(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Transactions of the Linnean Society of London, 2nd series? Botany"

THE 



TEANSACTIONS 



Of 



THE LINNEA^ SOCIETY 



OF 



LONDON. 



SECOND SERIES— VOLUME IV. 
BOTANY. 




LONDON: 

PRINTED BT TAYLOR A^"D FRA^"CIS, RED LIOX COURI, FLTi:ET STRKi:T: 

SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON-HOUSE; 
AND BY LONGMANS, GREEN, AND CO., PATERNOSTER-ROW. 



1891-96. 

Mo.Bot. Guisicu 
J398. 



mm 



CONTENTS. 



PAET I.— May, 1894. 

I. The Flants of Milanji, Nyasa-land, collected by Mr. Alexander Whyte, F.L.S., 

and described by Messrs. Britten, E. G. Baker, Bendle, Gepp, and others; 
with an Introduction by William Carruthers, F.B.S., F.L.S. (Plates I.-X., 
and Map.) page 1 

PART II.— December, 1894. 

II. On the Flora of Mount Kinabalu, in North Borneo. By 0. Stapf, Ph.D., 

Assistant for Bidia, Boyal Herbarium., Kew. {Communicated by W. T. Thiselton 
Dye^, C.M.G., F.B.S., FL.S.). (Plates XI.-XX.) 69 

PABT III.— December, 1895. 

III. The Phanerogamic Botany of the Matto Grosso Fxpedition, 1891-92. By 

Spencer Le M. Moore, B.Sc, F.L.S., Botanist to the Expedition. (Plates 
XXI.-XXXIX., and Map.) 265 

PABT IV.— March, 1896. 
Titlepage, Contents, and Index 517 



ERRATA. 



^ag 


e 19, 


line 23, 


for 


Hohnelii 


»j 


83, 


51 


17, 




platyplujlla 


») 


87, 


11 


22, 




nervifoUa 


»i 


96, 


11 


22, 




Dichotrichium 


j> 


98, 


>1 


IG, 




ampullacea 


i» 


103, 


11 


li 




hracteata 


'? 


117, 


11 


13, 




aborginal 


" 


120, 


11 


13, 




platyphylla 




11 


11 


42, 




lasioclada 




121, 


U 


15, 




aurantiacum 




ij 


51 


49, 




Chirifa areolata 




ff 


)f 


52, 




galeopsis 




122, 


V 


25, 




Kingii 




155, 


>» 


4, 




Iluvilfindii 




204, 


)1 


*K 




crenata 




211, 


11 


29, 




incarui 




222, 


11 


26, 




Ilavllaudii 




226, 


1» 


6, 




SETIFLOllUS 




227, 


)J 


10, 




erenata 




237, 


51 


4, 




ScORTEcniNII, Shlpf, 




5» 


51 


5, 




major, Ridl., 




249, 


55 


40, 




pruniata 




262, 


55 


26, 




pendula 




26^. 




25, 




asperiplium 




2*;!), 


., 


25, 




Ichnosijjha 




270, 


15 


24, 




njdoniafolia 




273, 


)5 


7, 




n,<e\h, 




276, 


55 


10, 




ThyaUta 




286, 


55 


28, 




(jlianthus 




290, 


51 


7, 




Salicea 



read lloehnelii. ■ 

„ platypetahi . 

„ neiiifolia. 

„ DicJiroti'ichuin. 

,, ampulla ria. 

„ neriifoUa. 

„ aboriginal. 

,. ])latypetala. 

V lasioclados. 

5. aurantiaca. 

„ Didymocarpiis arrolatU" 

,, GalcopsiH. 

,. major. 

,5 cordatus. 

„ cnnnldta. 

„ al/ncans. 

,. «fihf(ciif(ls. 

1, .SKTIFLOKA. 

„ crenulnta. 
„ MAJOR, Hidl. 

Scortechiiiii, Sfapf. 

pruinata. 

pendulijtora. 

,, anperifoUviii . 

1, Jschnosiphoii. 

,, cifdoniti-folia. 

„ Jtuellia, 

Thryallis. 

., CyhiantJms. 

,, Sahicea. 






2nd Ser. BOTANY.] 



[VOL. IV. PAET 1. 



THE 



rii 



TEANSACTIO^S 



OF 



THE LINNEAK SOCIETY OF LONDON. 



THE PLANTS OF MILANJI. NYASA-LANJ), 

COLLECTED BY Mr. ALEXANDER WHYTE, F.L.S., 

AM) DE8G1UBED BY 

Messrs. BEITTEX, E. G. BAKER, REX DEE, GEFP, AXD OTHERS 
TFTTJI AX INTRODVCTION 

BY 

WILLIAM CARRUTHERS. E.R.S.. E.RS. 




L O X D O X : 

PllI^TEl) FOR raE i.IXNEAX SOCIETY 

:, . .-;,,>a \K;i FSANC!^, aSO LION GOUaT, FLEET STK£F!T.. 

SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTOS-HOIJSE. PICCADILLY, W 
AND BY I.ONGMA'XS, GREBJi> AXT> CO., PATERNO?4TEIl-flOW. 



Tkans. Lias. Soc, Ser. 2, Uot. Vol. IV.j 



[2'oface p. 1. 



•i4- 




/:j;»iJier. 



MAP TO ILLUSTEATE Mil. A. WHYTE'S BOTANICAL COLLECTIONS. 

lOriginaJhj prepared to illustrate Mr. H. H. Johnston's r^mari-^ on the Zoology 

(Proc. Zool. Soc. 1893, p. 617), and reissued by permission.] 



TRANSACTIONS 



OF 



THE LINNEAN SOCIETY. 



I. The Plants of Milanji, Nyasa-landj collected by 31r. Al&itand^r IFhyte, F.L.S., atid 
described by Messrs. Britte!^, E. G. J^ake^^ Rexdle, Gepp, and others ; with an 
Introduction by William Careuth^s, F.B.S., F.L.S. 

(Plates I.-X.) 

Kead 19th January, 1893. 

Introductiok. 

At the instigation of Mr. H. H. Johnston, C.B., Her Majesty's Commissioner and Consul- 
General in British Central Africa, Mr. Whyte, in October and November 1891, explored 
the Natural History of the mountain and district of Milanji, a region in Nyasa-land to the 
south of Lake Shirwa, in about 16° South latitude and 35 J° East longitude. 

Erom Mr. AYhyte's Report to Commissioner Johnston we learn that Milanji is an 
isolated range of, for the greater part, precipitous mountains, the main mass forming a 
huge natural fortress of weather-worn precipices, or very steep rocky ascents, sparsely 
clothed with vegetation. Many of its gullies and ravines are well wooded, and in some 
of them fine examples of grand African virgin forest are met with. The route by which 
Mr. Whyte ascended the mountain on the 20th October led up its south-east face, and 
at first zigzagged over steep grassy hills, do^vn precipitous gorges, and across rocky 
streams with beds of large water-worn granite boulders, which, when flooded, become 
impassable mountain-torrents. Eurther on the ascent became more diflB.cult, and he 

SECOXD series. — BOTANY, VOL. IV. B 



2 THE PLANTS OF MILANJI, NYASA-LAND, 

clambered over precipices, holding on by tufts of grass and scrub, which gave but 
slender support and scanty foothold. Once round these precipitous bluffs an interesting 
wooded gorge was entered, still steep and difficult, but with better foothold on the 
projecting rocks and tree-roots ; and most welcome was the kindly shade after hours of 
toil in a burning sun, rendered doubly fierce by the reflection from the scorching hot 
rocks. 

An interesting change in the vegetation was perceptible, plants of the lower slopes 
being mostly replaced by species new to Mr. Whyte, and in many instances approaching 
the flora of temperate climes, such as brambles and well-known forms of papilionaceous 
and composite plants. Perns, too, became more numerous, and now and again he 
scrambled through fairy dells of mosses, ferns, selaginellas, and balsams, with miniature 
water-falls showering their life-giving spray on the little verdant glades, while overhead 
hoary lichens and bright festoons of elegant long-tasselled Jycopods hung from the 
moss-covered ancient-looking trees. Up and up he climbed the apparently endless 
ladder of roots and rocks. Then he passed through a dense thicket of bamboo, and 
again found himself confronted by an ugly barrier of precipitous cliffs, which were duly 
surmounted with the friendly aid of tufts of a tussock-grass springing from the crevices 
of the rocks. Another hour's climb up a steep grassy glen brought him to the crest of 
the highest ridge. 

Here the scene spread out to view, and the climate, were such as fully to repay the 
explorer for a day of weary toil. Looking westward, he saw mapped out beneath him 
the plateau or basin of Milanji, with its rolling hills of grassy sward, its clearly defined 
belts of dark-green forest, and its numerous ravines and rivulets, all shaping their 
course towards the principal valley of the plateau, through which the Lutshenya'' flows. 
The climate was delightfully cool and bracing. Daring the forenoon, on the lower ridges 
of the mountains, at over 4000 feet lower than this point, he had sweltered in a stiflhii? 
heat of 106 degrees Fahrenheit in the shade ; while here he revelled in a clear, dry"^ 
health-restoring atmosphere of 60 degrees Eahrenheit. From this ridge, which forms 
one of the amphitheatre of hills surrounding tlie plateau or crater-like basin of Milanji, 
a good idea of the mountain-system is gained. Still looking towards the west, one sees 
on the right hand the main peaks of the mountain, rising directly from the valley of the 
Lutshenya, which runs parallel with its southern base, the height of one of the two 
summit-peaks having been calculated at 9,300 feet above sea-level. Across the table- 
land, in the distance, is the somewhat isolated and precipitous Tsharabi Mountain, which, 
with its own smaller plateau, is separated from Milanji tableland by the rocky valley 
and gorge of the Likabula River. To the front and to the left hand, there is a contin- 
uation of the rolling and grassy hills which encircle the plateau, and which are capped 
with rugged cliffs of scarped granite and gneiss rocks. 

Mr. Whyte spent two weeks on the plateau, changing to three different sites each 
distant from 5 to 7 miles from the other, and which enabled him to explore 'more 
thoroughly this new and interesting mountain-country. Unfortunately the rains and 
mists set in before he left, and consequently he had only nine good collecting days 



COLLECTED BY ME. ALEXANDER WHYTE. 3 

The flora of the mountain proved to be most interesting, the species met with being 
mostly distinct from those of the plain or even the lower slopes. Tlie most strikin'^ 
botanical feature of the plateau is a large cypress. The remnant left of this fine conifer 
is confined to a few of the upper ravines and valleys, the largest forest of them findiui^a 
comparatively secure habitat in the damp gorges of the Lutshenya valley. A few old 
scorched monarchs of the glen lead a precarious existence pretty well up the soutliern 
slopes of the main mountain ; but, unless steps are taken to protect them, these interestmg 
relics of the past are doomed to speedy destruction. It is deplorable to witness the 
devastating effects of the annual bush-fires, from which even this lofty and all but 
inaccessible retreat is not exempt. During the dry months of August and September 
these fires, originating from the villages on the lower slopes of the mountain, gradually 
creep up the precipitous cliffs from tuft to tuft of grass until at last they reach the 
grassy plateau. Once there, the work of destruction is rapid. The fire rages over the 
table-land and eats its way along the edges of the remaining belts of forest, annually 
scorching, if not burning, the bark and timber of the outside trees, and killing outright 
the young seedlings. In exceptionally dry seasons it appears that these fires have even 
penetrated some of the damp forests, and hundreds of giant cypresses lay prostrate and 
piled on each other in all stages of destruction, but generally burnt right through at 
the base of the tree. Mr. Wliyte measured several of these dead conifers, and one (by no 
means the largest met with) was 140 feet in length and 5 J feet in diameter at feet from 
its base, with a clear straight stem of 90 feet. The cones of this species of cypress 
(which may be new to science) are somewhat smaller than a chestnut, and open into four 
scales, each having a spur at its apex, and covering five or six winged seeds. The foliage 
is juniper-like, and the timber is of a dull reddish white colour, of excellent quality and 
easily worked. The bark on old trees is of great thickness. 

Tree-ferns attain a great size in the damp shady forests of the plateau ; one measured 
was 30 feet in height and 2 feet in diameter at its base. 

Mr. Whyte was impressed with the gorgeous displays of wild flowers in some favoured 
nooks of these highlands. There he observed creamy-white and yellow helichrysums, 
mingling with purple aud blue orchids and irises, and graceful snow-white anemones, all 
blooming in wild profusion, and rearing their heads from a bed of bright-green grassy 
sward — a floral carpet w^hich Nature alone can fashion. Altogether he procured several 
thousand specimens of dried plants. The grass-lands had been too recently burned to 
permit many of the plants reaching their flowering stage. 

Of the 64 Natural Orders to which the flowering plants belong, nearly half are 
represented by single species. Comjiositce are most numerous, having 50 species, and 
are followed by Legumhiosce with 28 species. 

The great majority of the known and described species obtained by Mr. Whyte from 
Milanji have already been noted from Tropical Africa. If to these are added the species 
here described for the first time, it appears that no less than 62 per cent, belong to 
tropical vegetation. Of the remainder a small number, amounting to 6 per cent., are 
plants widely distributed over the tropical and subtropical regions of the world, such as 

B 2 



4 THE PLANTS OF MILANJT, NTASA-LAND, 

species of Stellaria, Ahutilon, Cassia, Ageratum, JBidens, and Celosia. Thirteen per 
cent, of the species are North- African plants, the majority being found in Abyssinia, 
while 19 per cent, are South African. Excluding the widely distributed tropical plants, 
only two species, one of Clausena and the other of Asystasia, occur in both North and 
South Africa. Milanji is obviously in the region where the floras of North and South 
Africa meet, and intermingle in nearly equal proportions with the tropical vegetation 
more proper to the region. 

Among the plants of interest which have a southern facies are two species of Erica, 
carrying this South-African genus into the Tropics, and the fine W^iddiHngtonia, which 
attains a great size, and is fitted to supply in the future a useful timber. Several 
Malagasy plants are in the collection, and among these there are species of JEpallage, 
Mascarenhasia, and JBrachysfephanus, genera that have hitherto been considered endemic 
to Madagascar *. 

It deserves to be noted that a large proportion of the Polypetalous plants belong to 
genera, and often to species, widely distributed over the face of the globe, while a large 
number of the Monopetalse are not known out of Africa. The actual state of the case is 
that 17 per cent, of the genera of the Polypetalse, 35 per cent, of the Monopetalse, 25 per 
cent, of the Apetalse, and 28 per cent, of the Monocotyledones are confined to Africa. 
In view of the greater antiquity of the less differentiated plants, it is remarkable that 
the Polypetalae, as tested by the genera included in this memoir, have secured the widest 
world-distribution. 

POLYPETAL^. (By Edmtjnd G. Baker, E.L.S.) 

Anemone Whytean A, Baker fil., n. sp. (Plate I. figs. 1-3.) Radix perennis; caule 
nuUo ; foliis longe petiolatis, biternatim sectis vel partitis, segmentis subcoriaceis 
ultimis cuneato-obovatis, serratis, apice apiculatis, marginibus plus minusve irre- 
gulariter acute lobatis ; petiolis canaliculatis, praecipue ad basin villosis ; scapis 
erectis, ramosis, 3-8-floris; involucri bracteis lanceolatis vel ovatis acutis, villosis, 
interdum trisectis, segmentis anguste ovatis vel lanceolatis; involucelli bracteis 
lanceolatis acutis ; pedicellis albo-pubescentibus ; sepalis albis, imbricatis, oblongis 
vel ovatis, externe pilosis ; stylis brevibus ; stigmatibus bifurcatis ; receptaculo 
ovoideo piloso; carpellis oblongis, numerosis, glabris vel glabriusculis, sine cauda 
plumosa. 

Sab. Milanji, alt. 6000 ft. No. 100. 

Scape 1-2 ft. high ; leaflets 1-1^ in. long, J-| in. broad ; sepals J-| in. long ; carpels 
y in. long. 

This is the second member of the genus which has been found in Tropical Africa, and 
I have called it after the discoverer, Mr. A. Whyte. It has somewhat the habit of a 
KnowUonia, but has not the green sepals which are characteristic of that genus. It 
differs from A. Thomsoni, Oliver, in Icones Plantarum, t. 1491, in the leaves being 

Since this was written four species of Brachystephanus have been described from Tropical Africa. 



COLLECTED BY ME. ALEXANDER WHYTE. 5 

biternate and not triternate, and it appears perfectly distinct from the three Cape and 
Natal species of Anemone, A. capensis, Linn., Bot. Mag. t. 716, A. caffray Eckl. & Zeyh., 
Harv. Thes. t. 7, and A. FatmimcBy Harv., Bot. Mag. t. C958. 

THALiCTRUii UHYXCHOCARPTJM, Dill, ct A. Rich, in Ann. Sc. Nat. s6r. II. xiv. p. 202 ; 
Oliver, Flora of Tropical Africa, i. p. 8. 
Hab. Milanji, alt. 6000 ft. No. 159. 
Distr. Mountains of Abyssinia ; Camaroons ; Clarence Peak, Fernando Po. 

CissAMPELOS Paeeira, Linn. ; Oliver, Fl. Trop. Afr. i. p. 15. 
Mob. Zomba. 
Widely spread in the Tropics. 

Oleome moxophylla, Linn. ; Oliver, I. c. i. p. 76 ; Trans. Linn. Soc. ser. I. xxix. t. 5. 
Sab. Zomba ; Milanji. No. 133. 
Widely spread in Tropical Africa. Occurs also at the Cape. 

Viola abyssixica, Steud. in Scliimp. PI. Abyss, ii. n. 983 ; Oliver, /. c. i. p. 105. 
Hab. Milanji, alt. 6000 ft. No. 79. 
Distr. Mts. of Abyssinia ; Camaroons ; Fernando Po ; Kilimanjaro ; Madagascar. 

Aphloia the^fohmis, Benn, PI. Jav. Uar. p. 192, in adnot., affinis. 
Bab. Milanji, alt. 6000 ft. 
Has rather larger flowers than the usual form of this plant. Malagasy affinity. 

PoLYGALA viRGATA, Thunb. ; Harv. in Harv. & Sond. Fl. Capensis, i. p. 85. 
Hab. Milanji, alt. 6000 ft. No. 101. 
JDistr. Cape, Natal. 

PoLYGALA PERSiCARiiEFOLiA, DC. Prod. i. p. 326 ; Oliver, I. c. i. p. 129. 
JELab. Zomba ; Milanji, alt. 6000 ft. No. 185. 
A very variable plant, widely spread in the Tropics of the Old World. 

Mtjraltia mixta, L. ; Harv. in Harv. & Sond. Fl. Capensis, i. p. 104. 
Sah. Milanji, alt. 6000-7000 ft. No. 1. 
Distr. Cape. 

Silene Burchellii, Otth, in DC. Prodr. i. p. 374 ; Oliver, I. c. i. p. 139. 

Hab. Milanji, alt. 6000 ft. No. 135. 

Distr. Abyssinia ; Cape ; also in Arabia and Syria acording to Oliver's * Flora of 
Tropical Africa,' but Kobrbach in bis monograph restricts it to Soutb Afiica. 



6 THE PLANTS OF MILANJI, NTASA-LAND, 

Ceeastitim afhicanum, Oliver, I. c. i. p. 141. 
Sab. Milanji. No. 115. 
round also on Mt. Kilimanjaro and the Camaroons. 

Stellaria media, Cyr. ; Smith, Eng. Bot. t. 537. 

Sab. Milanji, 6000 ft. No. 147. 

Differs from the plant as found in this country by its much thicker leaves and more 
congested inflorescence. 

Hypericum lanceolatum, Lam. ; Oliver, PL Trop. Afr. i. p. 156. 

Sab. Milanji, 6000 ft. 

Sistr. Abyssinia ; Camaroons ; Fernando Po ; Kilimanjaro ; Masai Highlands ; 
Manganya Hills. Previously collected in Shire Highlands by Last and Buchanan. 
Occurs also in Madagascar. 

Hypericum peplidieolium, A. Bich., var. robustum, Baker fil. Caulibus adscend- 
entibus; foliis glabris oblongis vel ovatis, quam ea typi majoribus; iloribus axil- 
laribus solitariis post anthesin deflexis. 
Sab. Zomba ; Milanji. No. 143. 

Abutilon indicum, Sweet, Hort. Brit. ed. i. p. 54 ; Oliver, L c. i. p. 186. 
Sab. Milanji, alt. 6000 ft. Nos. 63, 154. 
Widely spread in the Tropics. 

Hibiscus cannabintjs, Linn. ; DC. Prod. i. p. 450 ; Oliver, I. c. I p. 204. 
Sab. Zomba. 
Widely spread in the Tropics. 

Hibiscus physaloides, Guill. & Perr. Fl. Seneg. p. 52 ; Harv. in Harv. & Sond. PI. 
Capensis, i. p. 172. 
Sab. Zomba ; Milanji. No. 196. 
Distr. Trop. Africa; Cape. J. M. Wood, no. 926, is this species. 

Triumfetta Mastersii, Baker fil., n. sp. (Triumfetta Wehoitsckii, Mast, in Oliver, 
Plora of Trop. Africa, i. p. 255, in parte.) Caule ligneo, ramis virgatis, canaliculatis', 
prsecipue superne cinereo-pubescentibus ; stipulis subulatis vel anguste lanceolatis', 
decidais ; foliis similibus Ceanothi foliis, breviter petiolatis, oblongis vel oblanceolatis' 
serratis, basi cuneatis, apice rotundatis vel subacutis, e caule florifero cum floribus 
co^taneis productis, discoloribus, palmatim 3-5-nervatis, supra Digrescentibus, subtus 
cmereo-pubeseentibus; floribus terminalibus ramoso-corymboso-cymosis ; allbastris 
externe fulvo-pubescentibus ; petaHs in sicco flavis, oblanceolatis vel anguste obovatis 



COLLECTED BY MR. ALEXANDER WHTTE. 7 

ad basin pubescentibus ; staminibus circiter 25 e toro elerato crassoque dispositis ; 
fructu cerasi parvi magnitudine, setis rectis ciliatis elongatis obtecto. 

Hab. Milanji, alt. 6000 ft. Nos. 103 and 105. 

Branches 1-2 ft. long ; leaves 3-3 J in. long, f in. broad. 

This plant is the same as a specimen collected by Sir J. Kirk on the Livingstone 
Expedition near Lake Nyasa. It appears distinct from the Angola plant, and I venture 
to separate the East-Tropical African species under the name Triumfetta Mastersii, and, 
to avoid confusion, append a short description of T. JVclicitschii*. 

T. Mastersii has also been collected in Nyasa-land by J. Buchanan, No. 13 1<1 ; it is 
closely allied to T. Rehmanni, Szyszyl. in * Rozprawy i sprawozdania z posiedzen,' p. 151 
(1887), from the Trans raal, and is also allied to the curious and distinct T. geoidea^ Welw. 

Geranittm simense, Hochst. in Schimp. PL Abyss, ii. p. 670 ; Oliver, I. c. i. p. 201. 
Sah. Milanji, alt. GOOO ft. No. 72. 
Bistr. Abyssinia ; Kilimanjaro ; Camaroons. Occurs also in Madagascar. 

Impatiens micrantha, Ilochst. ? in Schimp, PI. Abyss, ii. n. 1151, affinis. 
Sab. Milanji, alt. 6000 ft. Nos. 15, 140. 
Distr. Abyssinia. Closely allied to /. capensis, Thunb. 

Impatiens shirexsis, Baker fil., n. sp. (Plate I. figs. 4-13.) Caule erecto, ligneo, 
ramoso ; foliis membranaceis, anguste ovatis acutis vel acuminatis, penninervatis, 
petiolatis, serratis, basi cuneatis; floribus axillaribus solitariis; pedunculis quam 
petiola longioribns vel subsequalibus, medio ebracteatis ; sepalis lateralibus parvis, 
anguste ovatis acutis vel subacuminatis ; labello naviculari abrupte calcarato, calcare 
brevi curvato (petalis sepalisquein sicco flavis; ; vexillo ad apicem mucronato, tam lato 
quam longo, alis bilobatis auriculatis ; ovario late linear! curvato ; stigmate capitato. 
Hab. Milanji, alt. 6000 ft. No. 55. 

Stem 1 ft. long, possibly more ; leaves 2J-4 in. long, |-1 J in. broad ; spur J in. long. 
The affinity of this plant is with Impatiens bicolor^ Hook, f., /. buccinalis, Hook, f., 
I. Mackeyana, Hook, f ., and the recently described /. Ehlersii, Schweinf., as the flowers 
are solitary and axillary, and the lip is boat-shaped and possesses a short incurved spur. 

* Triumfetta Welwitschii, Mast, in Oliver, Fl. Trop. Afr. i. p. 255, in parte. Caule ligneo, ramis virgatis tenuiter 
canaliculatis, plus minusve teretibus, superne cinereo-pubescentibus ; stipulis subulatis subpersistentibus, petiolis 
longioribus, foliis breviter petiolatis, longioribus et angustioribus quam T. Mastersii folia, basi cuneatis, apice 
acutis, tenuiter serratis, e caule florifero post anthesin productis, parce discoloribus, penninervatis, interdum 
subpalmatinervatis ; floribus terminalibus ramoso-corymboso-cymosis, alabastris externa fulvo-tomentosis, petalis 
luteis, obovatis vel oblanceolatis ; fructu mihi ignoto, ex descriptione Welwitschii fructui T. Mastersii simili. 
Hah. Angola ; Pungo Andongo, Welwitsch. 
Branches 1-2 ft, long ; leaves 4-5 in. long, ^ in. broad. 

T. Welwitschii diifers from T. Mastersii in having longer and narrower leaves, which are produced on the same 
stem as, but subsequently to, the flowers. 



8 THE PLANTS OF MILANJI, NTASA-LAND, 

OxALis SENSiTiVA, Linn. ; Oliver, Z. c. 1. p. 297. 
Sab. Zomba. 
Widely distributed in the Tropics. 

Claijsena in-^qualis, Benth. in Hook. Niger Fl. p. 257 ; Oliver, PI. Trop. Afr. i. 
p. 307. 
Kab. Milanji. No. 49. 
Distr. Abyssinia ; Camaroons ; Cape. 

Gtmnosporia laurina, Szyszyl. in * Hozprawy i sprawozdania z posiedzen,' xviii. p. 35, 
forma. 

Scytophyllum laurinum, Eckl. & Zeyh. ; Soud. in Harv. & Sond. Fl. Capensis, i. p. 471. 
Celasirus laurinus, Thunb, Fl. Capensis, p. 217. 

Hab. Milanji, alt. 6000 ft. No. 51. 

The specimens from Milanji differ slightly in the shape of the leaves from the type 
found at the Cape of Good Hope. 

There is another species of Gymnosporia in the collection, No. 50, but the material is 
not sufficient to settle the species. 

Phtlica spicata, Linn. ? ; Harv. in Harv. & Sond. PL Capensis, i. p. 491. 
Sab. Milanji, alt. 6000 ft. No. 41. 
Distr. Cape. 

ViTis SERPENS, Baker, in Oliver, PI. Trop. Afr. i. p. 412. 
Cissus serpens, Hochst. in A. Rich. Fl. Abyss, i. p. 111. 
Sab. Zomba. 

Distr. Abyssinia ; Senegambia. 
There is another species of Vitis, but the material is insufficient for determination. 

Paullinia pinnata, Linn. ; Baker, in Oliver, PL Trop. Afr. i. p. 419. 
Sab. Zomba. 
Also found in Tropical America. 

Ceotalaria natalitia, Meissn. in Hook. Lond. Journ. Bot. ii. p. 67 ; Baker, I. c. ii. 
p. 34. 
Sab. Zomba. 
Distr. Natal ; Lower Guinea. 

Crotalaria recta, Steud. ; Baker, I. c. ii. p. 40. 
Sab. Milanji. 
Distr, Abyssinia; Zambesi-land. 



COLLECTED BT MR. ALEXANDER WIITTE. 9 

ADENOCAKPrs Mannii, Hook. fil. in Journ. Linn. Soc. vi, p. 8. 
Eab. Milanji, alt. 6000 ft. Nos. 57, 127. 
Collected also by J. Buchanan in Nyasa-land, No. 988. 
Diatr. Kilimanjaro; Camaroons; Fernando Po. 

Lotus arabictjs, Linn. ; Baker, in Oliver, Fl. Trop. Afr. ii. p. 62. 

Sab. Milanji, alt. 6000 ft. Nos. 80, 142. 

Bistr. Abyssinia ; Senegambia ; Zambesi-land. Occurs also in the Eastern Mediter- 
ranean Region and Arabia. 

iNDiGorERA Lyallii, Baker, in Journ. linn. Soc. xx. p. 128. 

Hab. Milanji. No. 11. 

A Mascarenc species, new to the Flora of Tropical Africa. Differs slightly from type 
in the pubescence on the underside of the leaflets. 

Tephrosia linearis, Pers. ; Baker, I. c, ii. p. 120. 
Kdb. Milanji. 
Distr. Nubia ; Upper and Lower Guinea. 

Tephrosia sericea, Baker, I. e. ii. p. 107. 
Hab. Zomba. 
Also collected on the Manganya Hills. 

Tephrosia White ana, Baker fil., n. sp. Fruticosa ; ramis superne petiolisque ferrugineo- 
pilosis ; foliis imparipinnatis, foliolis oblongis, apice mucronatis, basi rotundatis, 
lateralibus 6-8-jugis, margine et subtus media vena ferrugineo-pubescente, stipulis 
lanceolatis vel ovatis acuminatis ; floribus racemosis, calycis tubo campanulato 
pubescente, lobis deltoideis, cuspidatis, duobus conniventibus ; petalis (in sicco roseis 
vel purpureo-roseis) vexillo rotundato unguiculato, alis inaequilateraliter obovatis, 
carinam subaequantibus ; staminibus monadelphis, filamento vexillari sublibero, 
stylo applanato, incurvato, pubescente ; ovario lineari-oblongo ; legumine recto 
lineari-oblongo, sessili, dense rufo-piloso, apice apiculato 6-8-spermo. 

JSab. Milanji. 

Leaflets l-lj in. long, nearly ^ in. broad, legume 1^-2 in. long. 

Near Tephrosia barbigera, Welw., which, however, is a robust annual. 

Tephrosia nyas^, Baker fil., n. sp. Caule ligneo ramoso, ramis junioribus albo- 
pubescentibus ; foliis pinnatim multifoliolatis, foliolis 6-10-jugis oblongis, apice 
apiculatis, basi rotundatis vel cuneatis, coriaceis, penninervatis, superne viridibus, 
parce pubescentibus, subtus argenteo-canescentibus, vena media ssepe ferrugineo- 
pubescente ; floribus ad apicem ramulorum aggregatis, bracteis ovatis acutis, 
pedicellis brevibus ; calycis tubo externe ferrugineo vel albo-ferrugineo, laciniis 
lanceolato-linearibus vel lanceolatis ; petalis (in sicco roseis vel purpureis) vexillis 

second series. — botany, vol. IV. 



10 THE PLANTS OF MILANJI, NTASA-LAND, 

suborbicularibus externe pubescentibus, carinse alls subsequilongis, stamine superiore 
ad medium libero ; legum^inibus junioribus dense ferrugineo-pilosis, circ. 5-spermis. 

Sab. Nyasa-land {J. Buchanan, No. 51). 

Leaflets 1-lf in. long, J to rather more than J in. broad ; standard about \ in. long. 

The flowers of this plant are borne in dense terminal heads at the end of the 
branchlets. 

JEscHTNOMENE siiFOLiA, Welw. ; Baker, in Oliver, PI. Trop. Afr. ii. p. 151. 

Sab. Zomba; Milanji. 

In the ' Plora of Tropical Africa ' the articulations of the fruit are described as solitary, 
which I find to be generally, but not always, the case. 

Previously collected by Welwitsch in Lower Guinea, Pungo Andongo, alt. 3800 ft. 

^SCHYNOMENE, Sp. 

Sab. Milanji, 6000 ft. 

Desmodium hietum, Guill. & Perr. Fl. Seneg. Tent. ii. p. 209 ; Baker, I. c. ii. p. 163. 
Sab. Zomba. 
Distr. Upper and Lower Guinea ; Zambesi-land ; Natal. 

Desmodium ascendens, DC. Prod. ii. p. 332 ; Baker, /. c. ii. p. 162. 

Sab. Zomba. 

A rather depauperated form of this common American species. It also occurs in 
Upper and Lower Guinea. 

MucuNA coriacea. Baker, I. c. ii. p. 187. 
Sab. Milanji. 
Distr. Zambesi-land, Manganya Hills. 

Canavalia obttjsieolia, DC. Prod. ii. p. 404 ; Baker, I. c. p. 190. 
Sab. Milanji. 
Widely distributed through the Tropics of both Hemispheres. 

ViGNA vexillata, Bcnth. in Mart. PI. Bras. xv. pt. 1, p. 193, t. 50. f. 1. 
Sab. Zomba ; Milanji. 
Widely dispersed through Tropical America. 

ViGNA LUTEOLA, Bcnth. l. c. p. 194, t. 50. f. 2. 
Sab. Zomba ; Milanji, alt. 6000 ft. 
Like the last, also found in Tropical America. 

DoLiCHOS ERECTUS, Baker fil., n. sp. Perennis, caule erecto, glabro vel glabriusculo, 
angulato vel subterete; foliis trifoliatis petiolatis, petiolis strictis patentibus vel 
erecto-patentibus, canaliculatis, foliolis ultimis sequilateralibus, ovatis vel ovato- 
lanceolatis acutis, apice breviter mucronatis, utrinque glabris vel glabriusculis, 
junioribus pubescentibus, basi rotundatis vel subcordatis, penninervatis, foliolis 



COLLECTED BY ME. ALEXANDEE WIITTE. 11 

lateralibus inaequilateralibus, stipulis lanceolatis acutis persistentibus ; pedunculis 
axillaribus, praecipue ultimis, longissimis, 3-8-floris ; floribus ad apices peduncu- 
lorum aggregatis, racemosis ; bracteis ovatis membranaceis ; scpalis brevibus ovatis ; 
petaKs (in sicco flavis) calyce multo longioribus; stylis geniculatis teretibus, 
stigmatibus clavatiformibus, ovariis linearibus. 
Hab. Zomba, MilaDJi, alt. 6000 ft. 

Stem 1-1 J ft., possibly more ; lower peduncles 5-7 in. long ; leaflets about 2 in. long , 
petals nearly J in. long. 

This plant is allied to one collected by Dr. Schweinfurth in Djurland, no. 1366. The 
latter has, however, much narrower leaflets. 

DolicJios erectus does not belong to the section Streptostylis of Welwitsch, the style 
being not twisted ; the erect habit and the few flowered racemes therefore make it easy 
of recognition. D, simplicifolius, recently figured in the ' Botanical Magazine ' (t. 7318), 
also from the Shir^ Highlands, has unifoliolate leaves. 

Cajanus indicus, Spreng. Syst. Veg. iii. p. 248 ; Baker, in Oliver, Fl. Trop. Afr. ii. p. 216. 
Mab. Milanji. No. 155. 
Cultivated throughout the Tropics. 

Eriosema cajanoides. Hook, fil., in Fl. Nigrit. p. 314 ; Baker, l. c. ii. p. 227. 
Sab. Milanji. 
Common in Tropical Africa ; extends southwards to Natal. 

Eriosema parviflorum, E. Mey. ; Baker, I. c. ii. p. 225. 
Sab. Milanji, alt. 6000 ft. No. 38. 
This may be an allied species, as the flowers are larger and not so depressed as usual. 

Eriosema shirensis, Baker fil., n. sp. (Plate II. figs. 1-4.) Caule erecto, basi ligneo, 
praecipue superne patentim albo- vel rufo-piloso ; foliolis 3 oblongo-lanceolatis, 
membranaceis, acutis, utrinque albo-pilosis, subtus vena media rufo-pilosa, foliolis 
lateralibus inaequilateralibus, petiolis brevibus, stipulis lanceolatis; pedunculis 
foliola subaequantibus, ad apicem floribus reflexis dense vel subdense racemosis ; 
sepalis lanceolatis acuminatis, albo-pilosis ; vexillo biauriculato, alis carina paullo 
longioribus vel subsequantibus ; stylo geniculato, ultra medium parce incrassato, 
apice capitato-stigmatoso ; legumine patente rufo-piloso, 2-spermo, oblique apiculato ; 
seminibus nigrescentibus, oblongis, curvatis. 
Sab, Zomba. 

Stem 8 in. to 1 ft. high ; leaflets 2J-4 in. long, J-J in. broad ; flowers J in. long ; 
legume J in. long, J in. broad. 

This plant is the same as Nos. 31 and 1350 collected by J. Buchanan in the Shir6 
Highlands. It comes near to Eriosema Burkei, Benth., and E. salignum, E. Mey., but 
differs from both in the pubescence. The flowering peduncle is clothed with rather long 
patent, white hairs. 

c2 



12 THE PLANTS OF IVaLANJ], NYASA-LAND, 

rLEMiNGiA MACROCALYX, Baker fil., n. sp. (Plate II. figs. 5-12.) Fruticosa, ramis 
flexuosis, dense f ulvo-tomentosis ; foliis trifoliatis, foliolis ovatis acutis, coriaceis vel 
subcoriaceis, basi rotundatis, petiolatis, pryecipue subtus molliter f ulvo-tomentosis, 
junioribus utrinque fulvo-tomentosis ; stipulis ovatis vel laaceolatis acutis, sericeo- 
pilosis ; racemis axillaribus 3-9-floris ; floribus breviter pedicellatis ; sepalis magnis 
lanceolatis acuminatis, anticis pauUo longioribus, quam corolla pauUo brevioribus ; 
vexillo obovato, parce bilobato, unguiculato (in sicco flavo, apice purpureo), alls 
oblongis (in sicco flavis), unguiculatis ; carina unguiculata (in sicco flava, apice 
purpurea) ; legumine sessili, subturgido, oblongo-apiculato, f ulvo-hirto. 
Bah. Milanji. No. 12. 

Leaflets J-IJ in. ; sepals J-f in. ; petals J in. long ; legume ^-f in. long, J in. 
broad. 

Diifers from JEriosema Jlemingioides, Baker, in Trans. Linn. Soc. ser. I. xxix. t. 34, to 
which it is allied, by its smaller leaves, much longer sepals, and fewer-flowered racemes. 

Pterocarpus melliferus, Welw. ; Baker, in Oliver, PI. Trop. Afr. ii. p. 239. 
JSab. Zomba. No. 133. 

*• Large forest-tree with profuse flowers like a Laburnum." — A, Whyte. 
Collected previously by Welwitsch in Lower Guinea, Pungo Andongo. 

Cassia mimosoides, Linn. ; Oliver, I. c. p. 280. 
Mab. Milanji. No. 130. 
Common in the Tropics of the Old "World. 

Cassia Petersiana, C. BoUe, in Peters, Mossamb., Bot. p. 13. 
Sab. Zomba. 
Bistr. Mozambique District. 

Brachtstegia globiflora, Benth., in Hook. Icones Plant. 1. 1359. 

Mab. Milanji or Zomba. 

The Shire Highlands seem to be the headquarters of the genus Brachystegia, as 
Buchanan has collected two species there, B. Jloribunda and B. longifoUa, and B. ap- 
pendiculata was collected at Zomba by the Livingstone Expedition 

Entada abtssinica, Steud. in Schimp. PI. Abyss, ii. p. 520. 
Hab. Zomba. 
Distr. Abyssinia; Angola. 

Albizzia fastigiata, E. Mey. ; Oliver, I. c. ii. p. 361. 
Bah. MUanji, alt. 6000 ft. 

Collected by J. Buchanan in Nyasa-land. No. 870. 
Bistr. Upper and Lower Guinea ; Natal. 



COLLECTED BY ME. ALEXANDER WHYTE. 13 

EiJBus APETALUS, Poir. ; Oliver, Fl. Trop. Afr. ii. p. 374. 
Sab. Milanji, alt. 6000 ft. No. 76. 
Distr. Abyssinia ; Mozambique District ; Madagascar ; Mascarene Islands. 

Etjbus huillensis, Welw. ; Oliver, I. c. ii. p. 376. 
Hab. Milanji. 
Distr. Angola. 

Alchemilla, sp. 

Sab. Milanji, alt. 6000 ft. No. 112. 

Cliffortia linearifolia, Eck. & Zeyh. ; Oliver, I. c. ii. p. 379. 
Sab. Milanji. 
Distr. Angola; Natiil; Cape. 

Choristtlis shirensis, Baker fil., n. sp. (Plate III. figs. 1-6.) Ramis ligneis, inferne 
glabris, superne minute puberulis ; foliis alternis, coriaceis, petiolatis, lanceolatis vel 
anguste ovatis acutis vel acuminatis, serratis, penninervatis, basi rotundatis vel late 
cuneatis, utrinque glabris; petiolis minute puberulis, quam lamina brcvioribus; 
floribus axillaribus dense paniculatis, paniculis foliis brcvioribus; calycis tubo 
obconico, lobis lanceolatis persistentibus ; petalis lanceolatis vel late subulatis, 
utrinque pubescentibus, sepalis paullo longioribus ; staminibus 5 ; stylis 2 brevibus, 
stigmatibus globosis, ovariis semi-inferioribus ; capsulis bilocularibus dehiscentibus, 
seminibus oblongis. 

Sab. Milanji, alt. 6000 ft. No. 53. 

Leaves 1^-2J in. long, f in. to nearly 1 in. broad. 

Collected also by J. Buchanan in the Shire Highlands, Nos. 158 and 1468. 

This is the second species of a formerly monotypic genus. Choristylis rhamnoidesy 
Harv. PI. Capensis, ii. p. 308, comes from Natal, and has broader and not such pointed 
leaves as the plant described above. 

TiLL^A AQTJATiCA, linn. ; Britten, in Oliver, PL Trop. Afr. ii. p. 388. 
Sab. Milanji. No. 95. 
Occurs also in North Europe and Germany. 

Orassula globtjlarioides, Britten, I. c. ii. p. 389. 
Sab. Milanji, alt. 6000 ft. No. 158. 
Collected previously by Dr. Meller on Mt. Chiradzura. 

Myrothamnus plabellifolia, Welw. ; Oliver, PI. Trop. Afr. ii, p. 404 ; Trans. Linn. 
Soc. ser. I. xxvii. p. 23, t. 8. 
Sab. Milanji, 6000 ft. 
Occurs also in Lower Guinea and at the Cape. 



14 THE PLANTS OF MILANJI, NYASA-LAND, 

PissoTis INCANA, Triana in Trans. Linn. Soc. ser. I. xxviii. p. 58 ; Cogn. Melastomacesey 
p. 370. 

Dissotis canescens, Hook: f., in Oliver^ Flora of Trop. Africa^ ii. p. 453. 
Osbeckia canescens, E. Mey. ; Graham in Bot. Mag. t. 3790. 

Sab. Milanji. No. 168. 

Distr. Nile land ; Natal ; Transvaal. 

PissoTis PEiNCEPS, Triana in Trans. Linn. Soc. ser. I. xxviii. p. 57 ; Cogn. Melastomaceae^ 
p. 375. 

Dissotis eximia, Hook, f., in Oliver, Flora of Trop. Africa, ii. p. 454. 
Sab. Milanji, 6000 ft. No. 103. Zomba. 
Distr. Central Africa ; Mozambique District ; Zambesi-land; Natal. 

PissoTis JoHNSTONiANA, Baker fil., n. sp. (Plate II. figs. 13-17.) Ramis robustis 
elongatis, plus minusve scabridis, ad nodos parce incrassatis, setosis, obscure tetragonis; 
foliis petiolatis, coriaceis, ovatis acutis vel acuminatis, basi rotundatis, utrinque 
scabridis, palmatim 5-nervatis, petiolis scabridis ; floribus terminalibus paniculatis ; 
pedicellis glabris ; calycis tubo angulato, sparse hirto, tamen glabro vel subglabro, 
lobis 5 ovatis vel oblongis, margine ciliatis, persistentibus; petalis magnis obovatis 
(in sicco purpureo-violaceis) ; staminibus insequalibus, connectivo magno ad basin 
producto ; stylis staminibus longioribus ; stigmate clavato curvato ; capsulis 5-valvis, 
ad apicem setis albis coronatis. 

Sab. Milanji, alt. 6000 ft. No. 74. 

Leaves 2-3 in. long, petioles J-| in., petals 1 in. long. 

The flowers of this plant are pentamerous and panicled. 

This plant has been named in honour of Commissioner H. H. Johnston, the organizer 
of the expedition. 

HoMALiTiM ArmcANTJM, Bcuth. in Journ. linn. Soc. iv. p. 35. 
Sab. Milanji. 
Distr. Upper Guinea. 

WoRMSKiOLDiA LGNGiPEDTJNcrLATA, Mast. in OUvcr, Tl. Trop. Afr. ii. p. 502. 
Sab. Milanji, alt. 3000-6000 ft. No. 97. 
Distr. Mozambique District. 

Tkyphostemma apetaltjm, Baker fil., n. sp. (Plate III. figs. 7-11.) Herbaceum, erectum, 
cirrhiferum, glaberrimum, glaucum ; foliis petiolatis, lanceolatis vel anguste lanceo- 
latis, apice mucronatis, serratis, ad basin laminae appendiculatis ; stipulis parvis, 
hnearibus, subulatis; appendiculis stipuliformibus 2 magnis, crescentiformibus, 
serratis, supra pedunculos e caule exorientibus ; ramulis floriferis gracilibus, strictis, 
erecto-patentibus, quam folia duplo brevioribus, 1-2-floris, supra florum insertionem 
subito attenuatis, demum iterum incrassatis, in cirrhum simplicem exeuntibus; 



COLLECTED BY ME. ALEXANDER WHTTE. 16 

pedicellis gracillimis ad basin appendiculatis ; sepalis 5 obtusis (in sicco flavo- 
viridibus) ; petalis nullis ; corona exteriore gamophylla, cyb'ndracea, ore longe 
fimbriata, corona interiore annulari brevissima ; staminibus inclusis ; stylis pra?cipuo 
post anthesin exsertis, divergentibus, filiformibus ; stigmatibus clavatiformibus ; 
fructu pericarpio chartaceo ovoideo ; seminibus fere ellipticis applanatis. 
Sab. Zomba. 

Stem 1-lJ ft. high ; leaves 2 in. to nearly 3 in. long, J in. to nearly \ in. broad ; 
peduncles f in. long. 

This plant resembles Tryphostemma zanzibaricum, Mast, in Oliver, PL Trop. Afr. ii. 
p. 508, in the structure of its flowers. Engler, in the ' Botanische Jahrbiicher ' for 1892, 
revises the genus Tryphostemma, proposing the sections Entry phostcnmia^ Neotrypho- 
stemma, and Basananthe, this last section, consisting of West-tropical African plants, 
having been formerly kept up as a genus. 

T. apetalwn will belong to the section Eutrypho8t€mma\ but the peduncles are 
1-2-flowered, and not 3-flowered. The stipuliform appendages referred to in the 
description are organs similar to those found in T. triloba y Bolus in Hook. Icones Plant, 
t. 1838, where, however, they are described as stipules ; their point of insertion is above 
the peduncle. 

MoMORDicA FCETiDA, Schum. ct Thonu. PI. Guin. p. 426. 
Sab. Zomba. 
Bistr. Abyssinia ; Cordofan ; Central Africa ; Guinea ; Fernando Po ; Natal. 

Htdrocottle asiatica, Linn. ; Hiem, in Oliver, PL Trop. Afr. iii. p. 6. 
Sah. Milanji, alt. 6000 ft. No. 197. 
Widely spread in Tropical and Subtropical regions. 

Phtsotrichia Buchanani, Benth. in Hook. Icones Plant, t. 13 58. 
Sab. Milanji. No. 14. 
Gathered on Zomba by Buchanan and Sir J. Kirk. 

Peucedanum, sp. 

Sab. Milanji, alt. 6000 ft. 

Caucalis melanantha, Benth. & Hook, fil. ; Hiem, in Oliver, Fl. Trop. Afr. iii. p. 26. 

Sab. Milanji. No. 121. 

Distr. Abyssinia, alt. 9300 ft. (Schimper) ; Camaroons, alt. 7000-9000 ft. , Kilima- 
njaro ; Masai Country ; Kapte Plateau, 5000-6000 ft. ; Plateau, Lake Nyasa ; also in 
Madagascar. 

Caucalis pedtjnculata. Baker fil., n. sp. C. radice crassa ; caule adscendente, canali- 
culato, albo-piloso; foliis ovatis, bipinnatifidis, segmentis oblongis acutis, incisis, 
inferioribus distincte petiolatis, vagina membranacea, venosa, magne dilatata, 
margine albo-pilosa ; umbellis terminalibus, involucri et involucelli bracteis lineari- 



16 THE PLANTS OF MILANJI, NYASA-LAND, 

lanceolatis ; floribus paryis (in sicco flavo-viridibus) ; petalis obovatis ; fructu flavo- 
viridi cum setis paulum pallidioribus, vittis primariis baud prominentibus setis 
breyibus armatis, vittis secondariis setis glocbidiatis longioribus armatis. 
Hab. Milanji. No. 160. 
Fruit ^ in. lon^. 

This plant was also collected at Blantyre by L. Scott, in Oct. 1887, and near Muata 
Menya, Zambesia, alt. 4000 ft., by Sir J. Kirk. 

Differs from C. melanantha, Benth. et Hook, fil., in having the secondary umbels 
borne on peduncles of half an inch in length and in the flowers being a different colour. 
C. melanantha has dark purple flowers. 

CUSSONIA, sp. 

Hah. Milanji, alt. 6000 ft. No. 54. 



MONOPETAL^. (By James Britten, F.L.S.) 

(AcANTHACE^, by S. Moore, E.L.S. ; Botena, by W. P. Hiern, M.A., F.L.S. ; 

Selago, by R. A. Rolfe, A.L.S.) 

Pentas * PURPTJiiEA, OHvcr, in Trans. Linn. Soc. ser. I. xxix. p. 83, vel aff. 
Sab. Milanji, 6000 ft. 

Otomeria dilatata, Hiern, in Oliver, Fl. Trop. Afr. iii. p. 60. 
Hab. Zomba ; Milanji, 6000 ft. No. 139. 

Pentodon decumbens, Hochst. in Flora, 1844, p. 552. 
Hab. Zomba. 

Oldenlandia (cf. O. caffra, Eckl. & Zeyh.). 
Hab. Milanji. 

Oldenlandia Oliveriana, K. Schum. in Engl. Hochgebirgsflora, p. 397. 
Hab. Milanji, 6000 ft. Nos. 126, 146, 177. 

PsTCHOTRiA HiRTELLA, Oliver, in Trans. Linn. Soc. ser. II. ii. p. 336. 
Hab. Milanji. 

Anthospermtjm Whtteanum, Britten, sp. n. Suffruticosum, virgatum ; caule erecto^ 
1-1^-pedali, in feme nudo, superne dense ramoso ; ramis erectis vel adscendentibus, 
brevibus, dense floriferis, ramis superioribus post anthesin ad 1^3 poll, productis ; 

* It may be worth while noting that Pentas mombagsana^ Hiern, mentioned by Prof. Oliver in Trans, Linn. Soc. 
ser. IL ii. p. 335, is a MS. name in the Kew Herbarium for the plant subsequently published by Mr. Hiern (Journ. 
Linn. Soc, xvi. p. 262) as P. parvifolia. 



COLLECTED BY MR. ^LEXAXDEK WHTTE. 17 

foliosis, saepe ramulosis ; foliis lauceolatis acutis, ^-^ poll, longis, vorticillatis vol 
fastii^iatis, marginibus revolutis, hirtis ; fructibus raiuiniis, l-3nis, obovato-cuneatis, 
birsutis, in axillis profuude iiisertis. Tlores nou vidi. 
Ilab. Milanji. No. 48. 

An erect woody undersliriib ; stems bare below, with numerous usually short branches 
above, densely covered with inflorescence in the lower portion and continued into barren 
leafy shoots lJ-3 in. 1od«^ ; the leaves and branches are clotlied w itli short hairs. 
Approaching A. emirnense in habit, but readily distinguished by the long terminal leafy 
shoots. 

Anthospermum lanceolatum, Thunb. Prod. PI. Cap. p. 32. 
Hah. Milanji. 
Widely distributed in South Africa. 

Spermacoce dtbr vchiata, Oliver, in Trans. Linn. Soc. ser. I. xxix. p. 88. 
Uab. Milanji, 0000 ft. Nos. IIA, 102. 

Galium Aparixe, Linn. Sp. PI. p. 108. 
Ilab. Milanji. No. 132. 
A mere fragment, probably of this species. 

Valeriana capensis, Thunb. Prod. PL Cap. p. 7. 

Uab. Milanji, 0000 ft. No. 111. 

The specimens differ from South African examples in the greater length of the 
elongated terminal leaflet, but 1 do not find other differences. The genus has hitherto 
been recorded from Tropical Africa only by Prof. Oliver (Trans. Linn. Soc. ser. II. ii. 
p. 337), who had a single flowering specimen, insufficient for description, from Kilima- 
njaro. 

Bothriocline Schimperi, Oliver, var. longipes, Oliver & Iliern, in PJ. Trop. Afr. iii. 
p. 200. 
Hab. Milanji, 0000 ft. Nos. 40, 125. 

Vernonia margixata, Oliver & Iliern, /. c. iii. p. 278. 
Hah. Zomba. 

Yerxonia pteropoda, Oliver & Iliern, /. c. iii. p. 283» 
Ilah. Milanji, 0000 ft. 

Verxoxia glabra, Yatke, in Oesterr. Bot. Zeitschr. xxvii. p. 194. 
Hab. Milanji, 0000 ft. 

Verxoxia Poskeaxa, Vatke & Hildebr. /. c, xxv. p. 324. 
Hah. Milanji, 0000 ft. Nos. 98, 179, 190. 

Verxoxia (§ Stexgelia) Whyteaxa, Britten, sp. n. Frutex erectus; ramis sub- 
angulosis, rigidis, tomentosis ; foliis 2-3 poll* longis, petiolatis, superioribus sub- 

SECOXD series. — BOTAXY, VOL. IV. D 



IB THE PLANTS OF MILAXJI, NYASA-LAND, 

sessilibus, cordato-ovatis acutis, irregulariter dentatis, superne puberulis, subtus 
cano-tonientosis, venis conspicuis ; petiolis 1 poll. long. ; pedunculis robustis ; 
capitulis panels (4-5) in cymis terminalibus corymbosis, late campanulatis, 1^-2 poll, 
diam. ; involucri squamis multiseriatis, laevibus, fusco-viridibus, late triangulatis, 
irregulariter crenato-dentatis, acutis, exterioribus recurvis, interioribus erectis 
lanceolatis, cornels, obtusatis, pappo proxime sequantibus ; pappo stramineo, aclienis 
fuscis. Corollani non vidi. 
Ilab. Zomba. 

A fine species, near V. adoeusis, Scb. Bip,, and V. dry mar ia^ Klatt. The specimens 
are all in fruit ; but in general appearance this species is very distinct from any that I 
have seen, and the shape of the leaves distinguishes it at once from its allies. 

Yernonia natalensts, Sch. ]3ip. in "VYalp. Eepert. ii. p. 947. 

Sab. Zomba and Milan ji {Whyte)\ Angola {Welwitsch, No. 3338). 

Vernonia natalensis, var. 
. Sab, Milan ji. 

Upper surface of the leaves green and nearly glabrous, underside silvery ; midrib 
conspicuous ; margin more or less revolute ; involucral scales short, broad, mucronate, 
with dark tips. Possibly a distinct species, between F". natalensis and V. 2nnifolia, 

Ageratum conyzoides, Linn. Sp. PL p. 839. 
Sab. Milanji. Ko. 187. 

EuPATORiUM AERiCANUM, Oliver & Hiern, in PI. Trop. Afr. iii. p. 301. 
Sah. Milanji. Nos. 66, 106 ; Zomba. 

DiCHRocEPHALA LATIFOLIA, PC. in Guill. Arcliiv. Bot. ii. p. 518. 
Sab. Milanji. 

NiDORELLA MiCROCEPHALA, Stectz, in Pcters, Mossamb., Bot. -p. 406. 
Sab. Milanji, No. 67 ; Zomba. 

CoNYZA VARIEGATA, Sch. Bip. cx A. Bich. Tent. PL Abyss, i. p. 388. 
Sab. Milanji. 

CoNYZA .EGYPTIACA, Ait. Hort. Kew, iii. p. 183. 
Sab. Milanji, 6000 ft. No. 153. 

Blumea lacera, DC. in Wight, Contrib. Bot. Ind. p. 14. 
Sab. Zomba. 

AcHYROCLiXE IIocHSTETTERi, Sch. Bip. ex A. Bich. Tent. PL Abyss, i. p. 429. 
Sab. Milanji, 6000 ft. No. 107. 

AciiYiiocLiNE ScHiMPERi, Sch. Bip. I. c. p. 428. 
Ilab. Milanji. 



COLLECTED BY ME. ALEXA>'DEli WHITE. 19 

Gnaphalium luteo-album, Linn. Sp. PI. j). Sol. 
Hah. Milanji. 

Helichrysum cymosum, D. Don in Mem. Wern. Soc. v. p. 550, var. X 
Eah. Milanji, GOOO ft. No. 32. 

Helichrysum nitens, Oliver & Hiern, in Fl. Trop. Afr. iii. p. 350. 
Hah. Milanji plateau, 7000 ft. No. 28. 

HELiCHRYSUii BucHANANi, Engl, llocligebirgsflora, p. 1;29. 

Huh. Milanji, 6000 ft. No. 134. 

" The natives resort to the plateau of Milanji period ieally to manufacture salt from 
this plant, which grows on the poor gravelly ground of the ridges on the plain," — 
A. Whyte. 

Helichrysum nudifolium, Less. Syn. Comp. p. 299. 
Kah. Milanji. 

Helichrysum Whyteanum, Britten, sp. n. Fruticosa, ramosa ; ramis lignosis, foliosis, 
alho-tomentosis ; foliis lineari-oblongis vel lanceolatis, coriaceis, acutis, aniplexi- 
caulibus, superne glabris, subtus albo-lanatis, marginibus recurvatis, ^-1 poll. 
long., -^i~\ poll. lat. ; capitulis late carapanulatis, solitariis vel 2-5nis, sessilibus 
vel breviter pedunculatis, argon teis, 1 poll, latis; involucri bracteis niultiseriatis, 
lanceolatis obtusis. 
Kah. Milanji plateau, GOOO-8000 ft. ( Whyte, No. 27) ; Nyasa-land {Buchanan, No. 

950) (1891). 

A handsome plant, near II. Neicii, Oliver & Hiern, in Fl. Trop. Afr. iii. p. 349, and 

H. Hohnelii, Schweinf. in Hohnel, Hudolj^h-See, p. 8G2. The young branches are leafy 

throughout, the older ones bare and almost glabrous below, with conspicuous leaf-scars. 

Helichrysum milanjiense, Britten, sp. n. Fruticosa, ramosa; ramis erectis, mono- 
cephalis ; foliis lineari-acutis, amplexicaulibus, adpressis, inferioribus recurvatis, 
superioribus patentibus, subtus laxe albo-lanatis, marginibus revolutis; capitulis 
campanulatis, sessilibus, luteis, J poll, longis, f poll, latis ; involucri bracteis ovato- 
lanceolatis acutis. 
Hab. Milanji plateau, 6000 ft. No. 52. 

Resembling H. Kirkii, but distinguished at once by the monocephalous branches all 
springing from near the root. Leaves about 1 inch long, closely adpre.ssed to the stem 
for a third of their length, the lower recurved ; lieads shining dull yellow. 

Helichrysum dexsiflorum, Oliver, in Hook. Ic. Plant, t. 2286. 

Eah. Milanji, 6000 ft. ( Whyte, No. 118) ; Nyasa-land {Buchanan, No. 933). 

Helichrysum latifolium, Less. Syn. Comp. p: 297. 
Eah. Milanji, 6000 ft. 

d2 



20 THE PLANTS OF MILANJI, NYASA-LAND, 

A Cape species, to which I refer the jDlant from Zambesi {Stewart) placed in El. Trop. 
Afr. iii. p. 353 as a form of S. gerhercefolimn. 

Helichrysum gerber^folium, Sch. Bip. ex A. Eich. Tent. PL Abyss, i, p. 425. 
Rah. Milanji, 6000 ft. No. 80. 

Helichrysum tjndatum, Less. Syn. Comp. p. 298. 
Hah. Zomba. 

Helichrysum atjrtculatum, Less. Svn. Comp. p. 311. 
Rah. Zomba. 

Athrixia rosmarinifolia, Oliver & Hiern, in PL Trop. Afr. iii. p. 355. 
Hah. Milanji, 6000 ft. No. 58. 

Epallage dextata, DC. Prod. vi. p. 4. 

Hah. Milanji, 6000 ft. No. 37. 

A common and variable Malagasy species ; genus not, I think, hitherto reported 
from the mainland. 

Blainvillea Gayana, Cass, in Diet. So. Nat. xlvii. p. 90. 
Hah. MilaDJi, 6000 ft. No. 94. 

Melaxthera abyssixica, Oliver & Hiern, in PL Trop. Afr. iii. p. 382. 
Hah. Zomba. 

Melaxthera Browxei, Sch. Bip. in Plora, 1844, p. 673. 
Hah. Zomba. 

Spilaxthes Acmella, Linn. Syst. Nat. ed. XIII. ii. p. 618. 
Hah. Milanji. 

BiDEXS piLOSA, Linn. Sp. PL 2^- 832. 
Hah. Milanji, 6000 ft. 

Gyxura cerxtja, Benth. in Hook. Niger Plora, p. 437. 
Hah. Zomba. 

Gyxura amplexicaulis, Oliver & Hiern, in PL Trop. Afr. iii. p. 403. 

JLil). ^filanji. 

Gyxura crepidioibes, Bentli. in Hook. Niger Plora, p. 438. 
Hah, Milanji, 6000 ft. No. 176. 

Ctxeharia kilimaxscharioa, Engl. Hochgebirgs flora, p. 439. 
Hah. Milanji, 6000 ft. No. 132. 



COLLECTED BY MR. ALEXANDER AVHYTE. 21 

To this is to be referred " Cineraria ahijssintca, Sch. Bip., forma," of ^Ii-. TT. Tf. 
Johnston's Kilima-njaro collections (Trans. Linn. Soc. ser. II. ii. p. 3 40). 

Sexecio latifolius, DC. Prod. vi. p. 3S7. {S. bnpleuroides, DC.) 
Ilab. Milanji, 600^ ft. Nos. 70, 109; Zoniba. 

Sexecio lasiorhizus, DC. ? I. c. 
Hab. Milanji. 

Senecio Whyteaxus, Britten, sp. n. Perennis, ij^laber, rliizoni^te obliquo; caule erecto. 
herbaceo, striato-sulcato, l-3pedali; foliis inferiorilms petiolatis, petiolo 1-2 poll, 
longo, decurrente, basi anijilexicauli, lamina lJ-3 poll, longa X 1 poll, lata, ovatis vel 
lanceolatis acutis, crenato-serratis, marline glandulosis; foliis superioribns lanceo- 
latis, subintegris, 1 poll, longis, bmcteiformibus ; panicula corymbosa, laxa ; pedicellis 
elongatis, patentibus, paucis; capitulis in quoquo pedicello solitariis suffidtis multi- 
floris, magnis, 1 poll, latis; bractoolis lineari-lanceolatis acutis; involucro niulti- 
bracteato ; bracteis lanceolatis acutis, 4-5 lin. longis, margine scarioso ; ligulis luteis, 
paucis, magnis, striatis, quam involucra duplo longioribus, 2 lin. latis. 
Hab. Milanji, GOOO ft. No. 101). 
Habit of S. laswrhizus^ DC, to which it bears considerable general resemblance. 

Leaves mostly from the rootstock or lower portion of the stem, passing into ntirrow 

almost entire bracts ; the teeth tipped with large cushion-like glands ; panicle lax, cni)itula 

few, solitary, on ascending peduncles, 3-6 inches long. 

Sexecio auriculatissimus, Britten, sp. n. Fruticosa, scandens, glabra ; foliis anrieulatis ; 
auriculis magnis, simplicibus, iutegris, ovato-cordatis, amplexicaulibus, ^-1 poll, 
longis; lamina 1-2 poll, lata, e basi cordata obreniformis, apice late depresso vel eroso, 
crenato-dentato, dentibus latis obtusis truncatis, subtus pallida, venis conspicuis; 
inflorescentia laxe corymbosa ; pedunculis bracteolatis, bracteolis subulatis ; florum 
marginalium tubus 2 lin. longus, ligula 3-4 lin. longa ; capitula late campanulata. 
Hab. Milanji, No. 83 ; also from Makua country {J. T. Last, 1887) and Shire Highlands 
[Buchanan, 1881). 

Well distinguished by the very large auricles, which are often cordate at the base, 
assuming a 2-lobed appearance, through which the petiole is continued ; the upper leaves 
are sometimes reduced to their auricles. I have not found ripe achenes, but Prof. Oliver 
places the plant in Senecio (Herb. Kew) ; it has the habit of some climbing Cinerarias. 

Othoxxa Whtteaxa, Britten, sp. n. (Plate IV. figs. 1, 2.) Pruticosa vel subfruticosa, 
erecta, glabra, multicaulis, pedalis, superne ramosissima; ramis crebris, brevibus ; foliis 
obovato-lanceolatis acutis, carnosis, l-lj poll, longis ; paniculis elongatis foliaceis, 
densis ; capitulis numerosis, campanulatis, nutantibus ; pedunculis bracteolatis, 
gracilibus, patentibus vel reflexis, post anthesin elongatis, erectis ; bracteis invo- 
lucri 5, lanceolatis acuminatis, J-1 poll, longis, margiuibus membranaceis ; papj)o 



22 THE PLANTS OF MILANJI, KTASA-LAND, 

post antliesin elon^ato, ereclo, jiatente, 1-1^^ poll, longo ; aclieniis 4 lin. longis, 2 lin, 
latis, costatis, glabris. 

Mab. Milanji, 6000 ft. No. G2. 

A very distinct species, with numerous erect wand-like stems, becoming leafless below, 
clothed above with obovate-lanceolate acute, fleshy, sessile leaves, tapering to the base, 
and short flowering branches, forming a dense, elongated, leafy panicle ; peduncles slender, 
spreading or reflexed during flowering, afterwards stouter and erect; heads campanulate, 
about ^ in. long wiiile flowering, and then nodding, afterwards erect, the upper opening 
first ; involucral bracts 5, lanceolate-acuminate, about v in. long during flowering, 
extending to 1 in. afterwards, with white membranous margins; pappus yellowish wliite, 
much elongated after flowering, l-l^J- in. long, spreading, resembling a shaving-brush; 
achenes smooth, or with a tew very short hairs. 

Allied to Othonna coriifolia (Sond.), with which O. BainesU, Oliver & Hiern, has been 
identified ; from which, however, it is very different in general appearance, as well as in 
the characters given above. It is the only species pecuhar to Tropical Africa. 

Bekkheya (§ Stob.i:a) Johnstoniana, Britten, sp. n. B. caule erecto, herbaceo, striate, 
irregulariter alato ; foliis radicalibus et inferioribus (5-7 poll, longis, 1 J latis) pinna- 
tifidis, subtus dense albo-tomentosis, supra viridibus dense setosis, setis luteis 
flaccidis, niarginibus setosis vel spinellosis ; petiolis decurrentibus ; foliis caulinis 
minoribus lineari-lanceolatis ; capitulis paucis (3-4), dense corymbosis, discoideis ; 
pedunculis brevibus, post antliesin elongatis ; squamis involucralibus lineari-lanceo- 
latis, serratis, spinescentibus acutis ; acheniis glabris, pappo pallidc fusco. 
Uab. Milanji. No. 7. 

Stem herbaceous, erect, conspicuously farrowed below, less so above, irregularly winged, 
radical and lower stem-leaves 5-7 in. long, 1 J broad, irregularly pinnatifid, clotted beneath 
with Avhite felty tomentum and above with weak yellowish spines, margins irregularly 
spinescent ; heads small, in dense corymbs of 3-4 peduncles, elongating after flowering ; 
flowers all discoid, purplish (?). The inflorescence, when young, resembles that of a 
Cirsium. 

Berkheya subulata, Harv. in Harv. & Sond. Tl. Capensis iii. p. 507. 
Mab. Zomba. 

ERYTHKOCEPirALrM ZAMBESiANiM, Oliver & Hiern, in PI. Trop. Afr. iii. p. 441. 
JIab. Milanji, 6000 ft.; Zomba. 

Gerbera abyssinica, Sch. Bip. in A. Bich. Tent. Fl. Abyss, i. p. 458. 
JIab. Milanji. No. 111. 

ToLPis ABYSSiNiCA, Sch. Bip. in Scbimp. PL Abyss, i. No. 1 r 
Ilab. Milanji; Zomba. 



COLLECTED BY MR. ALEXANDER WiiYTE. 23 

CllEPIS, sp. 

Hab. Milanji, 6000 ft. No. 111. 

Lactuca capensis, Thunb. Prod. PL Cap. p. 139. 
Hab. Milanji ; Zoniba. 

Lobelia tervexs, Thunb. Fl. Cap. ii. p. 40? 
Hab. Milanji. No. 100. 

Lobelia, sp. 

Sab. Milanji, GOOO ft. Nos. 21, 182. 

Wahlenbeegia oppositifolia, a. DC. Prod, vii. p. 429. 
Hab. Milanji. 

Wahlenbergia, sp. 

Hab, Milanji, GOOO ft. No. 101. 

LiGiiTFOOTiA ABYSSiNiCA, Ilochst. ex A. Rich. Tent. Fl. Abyss, ii. p. 1. 
Hab. Milanji, No. 117; Zoniba. 



Vaccinium AFRiCAxrM, Britten, sp. n. (Plate IV. fip^s. IJ-5.) Frutex vcl arbor parva; 
ramis glabris, erectis vcl adscendentibus ; foliis sessilibvis vol brevitcr pedunculatis, 
glabris, ovatis acutis, serratis, subtus reticulatis, 2-3 poll, longis, J-f poll, latis ; 
racemis bracteatis, axillaribus 1-2 poll, longis ; floribiis pedunculatis, solitariis vel 
binis, nutantibus, subglobosis, calycis segmentis triangularibus, glabris, patentibus 
acutis ; corollae subcampanulatse, 5-fida?, segmentis brevibus recurvatis ; starainibus 
corollani cequantibus vel pauUo longioribus, Fructus non vidi. 
Hab. Milanji, 7000-8000 ft. 
This is, I believe, the first Vaccinium recorded from the African continent. Mr. Whyte 

collected it in considerable abundance, and its very numerous (white or red ?) flowers 

doubtless render it a conspicuous figure in the landscape. 

Erica Johis^stoniana, Britten, sp. n. (Plate Y. figs. 1-6.) Fruticulus adscendens vel 
erectus; folia 4-na ovata, 3-4 lin. longa, 1-1 J lata, dense imbricata, adscendentia, 
incurva, marginibus revolutis pectinatis ; flores dense capitati, axillares, nutantes, 
breviter pedicellati ; sepala lineari-lanceolata, marginibus dense hirsutis ; corolla 
calycem aequans vel paulo excedens, campanulata ; antheris appendiculatis, appen- 
diculis tomentosis semi-orbicularibus, marginibus fimbriatis. 
Hab. Milanji plateau, 6000-6500 ft. No. 4. 

A low, much-branched shrub ; branches especially numerous on the upper part of the 
stem above the flowers, erect, of a woolly appearance owing to the close aggregation at 
the tips of the white hairs which fringe the leaves ; flowers red ?, 4-12 together in dense, 
nodding, globose heads on short axillary branches. It seems to come nearest to 
B. Solandra, Andr. 



24 THE PLANTS OF MILANJI, NYASA-LAND, 

Erica Whyteana, Eritten, sp. n. (Plate YI. figs. 7-12.) Eruticulus erectus, glaber- 
rinius, ramosus ; ramulis paucis, erectis vel adscendentibus, junioribus dense fob'osis ; 
foliis 4-nis, superne aggregatis, lanceolatis, rigidis, aciculatis, 4 lin. longis, J lin. latis ; 
inflorescentia racemosa, racemis bracteatis, axillaribus paucifloris vel terminalibus, 
multifloris ; bracteis foliaceis ; floribus breviter pedicellatis, sepalis patentibus, acumi- 
natis, 1 lin. long. ; corolla urceolata, persistente, 2 lin. longa ; staminibus membra- 
naceis, inclusis, antheris magnis appendiculatis, appendiculis parvis lanceolatis. 
Hah. Milanji, 6000 ft. No. 59. 

A smooth plant of virgate habit, the older branches leafless, or nearly so, for more 
than half their length ; the flowers when dry are of a brittle texture ; they appear to be 
orange or red. 

"With the exception of the widely distributed E. arhorea^ I am not aware that this 
characteristic Cape genus has hitherto been recorded from Tropical Africa. 

Philippia milanjiensis, Britten & Rendle, sp. n. Fruticulus ramulis pubescentibus ; 
foliis 4-nis, crassiusculis, oblongo-linearibus, glabris, suberectis, breviter petiolatis ; 
floribus parvis, ad apices ramulorum pluribus umbellatis glomeratis, breviter 
pedicellatis ; calycis lobis 4 quam corolla brevioribus, subrotundis, unius apice 
elongato, crassiuscnlo ; corollae brevis campanulatae lobis 4 rotundis, rubris ; 
antheris valde exsertis, liberis ; ovario subgloboso, albo-incano ; stigmate magno, 
peltato. 
A much-branched shrub, the slender ultimate branches finely pubescent ; the leaves are 
arranged in fours, deciduous, glabrous, and 1-1 J lines long. One of the sepals is produced 
into a short, rather thick tail ; the reddish corolla is \ line long ; the strongly exserted 
stamens are f line long, the anther \ line. The ovary is \ line long, subglobose, and 
covered with short white hairs. The stigma is large and capitate. 
Mah. Milanji, 6000-8000 ft. 

Philippia benguellensis, Welw. in herb. (Salaxls benguellensis, Engl. Hochgebirgs- 
flora, p. 328.) 

Bab. Milanji, 7000 ft. No. 31. 

I do not know why Prof. Engler removes this from Philippia, in which genus 
Welvvitsch, after careful examination, had placed it. In the absence of ripe fruit, it is 
impossible to say in which genus the plant should rest, but the general facies and the 
distribution of the genus induce me to leave it in Philippia. The clavis character in 
Gen. Plant. — " Stamina 4 " — would exclude it from Salaxis, but in the full description 
of that genus we read " Stamina 6-8." 

Welwitsch notes that this is called by the Portuguese colonists " Cedro pequeno " or 
" Cedro pequinino." Mr. Whyte has a specimen which is only in bud, but seems to 
be a new Philippia, which he labels as the foliage of a Cypress. 

Blaeria setulosa, Welw. ex Engl. Hochgebirgsflora, p. 328. 
Ilab. Milanji plateau, 6000-7000 ft. No. 26. 



COLLECTED BY ME. ALEXAXDET? WHYTK. 25 

KoTENA Whyteaxa, Hicm, sp. n. Foliis lanccolatis vel angustc ovali-ovatis, tcnuittM* 
coriaceis, margino soriceo-sotoso, hasi subcordatis, brovit(T potiolatis; florilnis 
solitariis, 6-7-mcris ; Ciilyco semi-fisso, lobis lanceolatis vel ovatis ; staminibus 8; 
ovario (in floribus niasculis rudimentario ?) piloso, in sty lis 2 dcsinento. 

Branches dark ashy; towards the extremities dark ferruginous and hispid, leafy. 
Leaves (of the young shoots) alternate, entire, lanceolate or narrowly oval-ovate, some- 
what cordate at base, acutely pointed at apex, thinly coriaceous ; silky, with pale tawny 
setose hairs on the margin and on the midrib both above and below; IJ-IJ in. long by 
|-f in. broad ; midrib depressed above ; petiole rj-^ in. long, clothed with tawny 
setose hairs. Flowers solitary. Pediificle axillary, about 4 in. long, clothed with 
tawny setose hairs, bracteate about the middle ; bnicts ()])posite or solitary, sessile, 
lanceolate, glabrous above, clothed beneath with tawny hairs, about J in. long. Calyx 
hemispherical, G-7-^eft, setose towards the base, J in. long ; teeth lanceolate or ovate, 
acute. Corolla f in. long, deeply divided ; lobes (J, glabrous, oblong or slightly obovate, 
obtusely pointed, net-veined, reflcxed. Stamens 8, i"V~5 i^* long; anthers narrowly 
linear-lanceolate, somewhat hairy at the back upwards, filaments very short. Ovary 
hairy, narrow (rudimentary ?) ; styles 2, glabrous. 

Moh, Milan ji. 

The affinity of the species is with J2. scabrida, ITarv. 

Tabern^montana ventricosa, Hochst. ex DC. Prod. viii. p. 366. 
Bab. Milanji. No. 188. 

TABERNiEMONTANA Stapfiana, Erittcu, sp. u. Foliis coriaccis oblongo-ellipticis, basi 
acutis, apice obtusis, 5-7 jjoll. longis, 2-3 poll, latis, nervis lateralibus prominenti- 
bus 12-14 ; floribus corymbose-paniculatis, paniculis in foliorum anni praecedentis 
axillis, pedunculo pedicellisquc crassis, calyce 5-lineo, corollae tubo pollicari sub- 
cylindrico, medio minus quam in T. angolensi ampliato, lobis latc-obovatis vel 
suborbicularibus, 16 lin. longis, 10-12 latis, apice rotundis, staminibus infra 
medium insertis fauces minime attingentibus, 5 lin. longis. 
Hab. Milanji, 6000 ft. No. 87. 

Differs from T. angolensis, Stapf (Kew Bulletin of Miscellaneous Information, 
1894, p. 23), in the more prominently-nerved leaves, thicker pedicels, and larger flowers. 
I have named this species after Dr. O. Stapf, to whom I am indebted for pointing out 
its affinity with his recently described T. angolensis. In that description he makes no 
reference to the fruit, which is absent both from the British Museum and Kew speci- 
mens. The plant is the " No. 55 (Apocynaceae) '* of the Apontamentoa (p. 588), where 
the fruit is described as " fructus foUicularis geminatus, mole, forma, et colore fruct. 
citr. aurantii. Semina membranis suberosis involuta (in optimo statu!)." To the 
specimen in the British Museum, where we have a complete transcript of all Welwitsch's 
notes, is attached a similar description, with the addition of '* Flores et fructus maturos 
et immaturos saepius eodem tempore fert." 

SECOND SERIES. — BOTANY, YOL. IV. E 



26 THE PLANTS OF MILANJI, NYASA-LAND, 

Mascaeenhasia variegata, Britten & Uendle, sp. n. (Plate VI. figs. 1-3.) Fratex 
gla"ber, ramosus ; foliis breviter petiolatis, ovalibus vel ovali-oblongis, subacuminatis 
vel obtusiusculis, rigide coriaceis, saepe variegatis ; floribus pedicellatis, interdum 
terminalibus et solitariis, interdum paucis in stirpibus brevibus axillaribus vel 
terminalibus fasciculatis ; calycis lobis ovatis, intus glanduliferis ; corollae liyper- 
crateriformis tubo supra medium, ad antherarum situm, dilatato, campanuliformi, 
lobis patentibus late ovatis, apice acuminatis, intus papillosis ; antheris sagittatis, 
filamentis brevissimis ; disci squamis planis, crassis, 4 per paria connatis, quinta 
libera, carpella distincta superantibus ; carpellorum parte superiore et stylo filiformi 
pilis fulvis indutis. 
The shoots are hard and woody ; the leaves vary in size from 3 J to 5 J in. long, and | to 
IJ in. broad, the larger ones are markedly variegated. The flowers are terminal and 
solitary, or 4-6 are crowded on a short terminal or axillary shoot, giving the appearance 
of an umbel ; the pedicel of the open flower is about J in. long. The calyx is 1 line 
long, and bears, inside at the base of the lobe, a ring of short, closely set, blunt, tooth-like 
glands. The corolla-tube is 5 lines long, and tapers slightly from below upwards to the 
middle, where it becomes dilated and campanulate ; at the dilated portion, which 
represents the insertion of the stamens, it is 1 J lines in diameter. The spreading corolla- 
lobes are 3^ lines long by 2J broad, and have a marked acuminate apex. The narrow 
sagittate anthers are 1| lines long. A divergence from the generic diagnosis in the 
* Genera Plantarum ' occurs in the fact that the disk-scales overtop the ovary. The latter 
is covered on the upper half wdth stiff yellowish hairs, and similar but shorter hairs clothe 
the thin style. Emit not seen. 

Agrees in many points with Baker's description of M. micrantha, but the leaves are 
much larger and narrow^er in proportion ; the flowers are also larger and have a much 
longer tube, that of M. micrantha being only twice the length of the calyx, which is 1 line 
long ; the limb of M. micrantha is only J in. across, that of M. variegata nearly | in. 

By the discovery of the present species, the genus Mascarenhasia, hitherto known 
only from Madagascar, is extended to the mainland. 
Sab. Milanji, 6000 ft., Oct. No. 108. 

EcTADiOPSis Welwitschii, Baill. in Bull. Soc. Linn. Paris, ii. p. 803. 
Hab. Milanji. No. 134. 

ScHizoGLOSSUM Nyas^, Britten Sc Bendle, sp. n. Suffrutex caule erecto, simplici, pubes- 
cente ; foliis sessilibus, cuneatis vel cuneato-oblongis, apice mucronulatis, utrinque 
glabris vel basi et marginibus revolutis, puberulis ; inflorescentia terminali umbellis 
pluribus lateralibus aucta ; bracteis subulatis, puberulis, quam florum pedicelli 
puberuli brevioribus; sepalis subulatis, sparse puberulis; petalis ovato-oblongis, 
multinerviis, dorso glabris, intus papillosis ; coronse lobis gynostegium superan- 
tibus planis, trilobatis, medio incrassatis, geminis appendiculis tenuibus suffultis ; 



COLLECTED BY ME. ALEXANDER WHTTE. 27 

gynostegio subsessili, coronoe lobis duplo breviore; styli apico pentagono, modio 

umbonato; antherarum connectivo apice membranaceo cxpanso, super stylum 

revoluto. 

The single specimen is 32 in. high ; the soft, woody, cylindrical, liollow stem 2^ lines 

in diameter for the lower 18 in., tapering gradually above in the flowering portion. 

The leaves are IJ-IJ in. long and 4-5 lines at tlic broiidest part, with a single prominent 

midrib and spreading lateral veins. The subulate flexible bracts are 2 lines long, the 

flower-pedicels 3-4 lines. The open flower measures nearly 5 lines across ; the sepals 

are If lines long, the petals 2J lines by 1 broad ; the corona-segments 1§ lines long, 

with a broad trilobed apex, the lateral lobes being short and rounded, and the central 

one more than twice as long and acute, and exceeding the acute subfalcate 

appendages. 

Hab. Milanji, Oct. 

SCHizoGLOSSUM BARBATUM, Britten & Rendle, sp. n. Sufli'utex caulibus e rhizomate 

crassiusculo ligneo pluribus tenuibus, sa^pissime simplicibus, pubtTulis ; foliis 

linearibus vel lineari-lanceolatis acutis, subglabris vel sparse puberulis, niargine 

revolutis ; inflorescentia umbellis pluribus, terminali; bracteis parvis, flliforniibus, 

deciduis ; florum pedicellis tenuibus, puberulis ; corolla) rotatae lobis anguste ovatLs, 

subacutis, reflexis, superne puberulis, dorso glabris ; sepalis subulatis, superne glabris, 

dorso et margine sparse puberulis ; corona) segmentis planis, intus medio bicarinatis, 

apice in caudam crassam, acuminatam, e flore exsertam productis ; gynostegio 

subsessili, styli apice piano pentagono ; antherarum connectivo velut in specie 

praecedente. 

The slender stems are 7-9 in. long, the lower leaves an inch or less in length, the 

upper 2-2 j in. The minute filiform bracts are 1 line long, the flower pedicels 3-4 lines. 

The open flower is 3-4 lines across; the sepals 1| lines long, the corolla-lobes 2 lines 

long by 1 line broad. The corona -segments are 2 lines long, the flexuose exserted tails 

often 1^ lines. 

This is at once distinguished from the preceding species by its thin stems, narrow leaves, 
and terminal inflorescence, and the shape of the corona-segments with their long and 
conspicuous tails. It seems near ^S*. datum, K. Schuni. (Engl. Bot. Jahrb. xvii. p. 123), 
judging from the description, but differs in having broader leaves, a terminal 
inflorescence, flowers twice the size, and very long corona-tails. 
Mab. MHanji, 6000 ft., Oct. 

STATHM0STEL3IA KEFLEXUM, Britten & Rcndle, sp. n.. (Plate VI. figs. 4-6.) Suffrutex 
caulibus erectis, strictis, super basin teretem complanatis, glabris, apice puberulis ; 
foliis subsessilibus, lineari-lanceolatis, elongatis, acutis, glabris, margine recurvatis, 
subscabridis ; inflorescentia umbeUata, terminali, simplici vel interdum ramo 
lateral! ; umbella 4-flora ; pedicellis tenuibus puberulis ; sepalis parvis, lanceolatis, 
acutis, dorso puberulis ; corollse abrupte reflexae, pallidae, lobis ovatis, tubo brevi ; 

E 2 



28 THE PLANTS OF MILANJI, NYASA-LAND, 

coronae lobis petaloideis, aurantiacis, g^'^nostegium alte superantibus, oblongis, 
cucuUatis, apice integris, medio dentibus binis productis et e cavitate appendiculam 
falcatam emittentibus ; styli apice orbiculari, 5-lobulato, medio depresso. 

The generally simple stems are 13-22 in. long ; the slender flaccid leaves have a midrib 
prominent on the lower surface and slightly puberulous, with no other venation visible ; 
the length varies considerably, from 2 J to 6 in., and the breadth from IJ to 2 J lines. 
The flow^er-pedicels are slender, 8-9 lines long ; the bracts have fallen. The flowers 
measure 4 lines across ; the small sepals (1 line long) are quite hidden by the corolla, 
which is abruptly reflexed almost at its base ; the short corolla-tube is f line long, the 
ovate lobes 3 lines by If broad. The large bright orange corona-segments are 3J lines 
long, and stand erect above the petals. The arms of the pollen-carriers gradually 
broaden from the attachment, forming a wide upper membranous portion with an 
incurving lower edge; the small pollinia-bearing portion is attached to this by an 
acuminate apex. 

The species is near S. incarnatum, K. Schum. in Engh Bot. Jahrb. xvii. p. 130, but 
distinguished by the oblong corona-lobe with its entire apex, that of S. incarnatum 
being truncate and subtrilobate ; the appendage is also much larger, projecting beyond 
the cavity, whereas in 8. incarnatum it is small and hidden in the cavity ; the arms of 
the pollen-carriers are also narrower in the Milanji species.* 

Hob, Milanji, Oct. 

* The following new species of this genus collected by Dr. Welwitsch, may be described here : — 
Stathmostelma Welwitschij, Britten & Rendle. SuiFrutex perennis, tubere dure napiformi, valde lactescens ; 
caulibus erectis, teretibus, glabris, ramosis ; foliis elongatis, lineari-lanceolatis, viridi-subglaucescentibus, 
subcarnosidis ; inflorescentia umbellata ; umbellis 2-6-floris ; pedicellis puberulis ; sepalis ovati-lanceolatis, 
acutis, subpubenilis; corollaa rotatse, intense aurantiacae, tubo perbrevi, lobis ovalibus acutis ; coronae lobis 
petaloideis, gynostegium valde superantibus, oblongis, apice truncatis, cucullatis, dentibus binis acutis lateralibus 
sub cucuUa instructis et medio appendiculam truncatam emittentibus, inferne carinatis ; gynostegio altiuscule 
pedicellate ; translatorum brachiis oblongis, basi inflexis, subito in filum tenue contractis, iterum in aream 
triangularem poUiniferam dilatatis. 
The plant is 2|-4 ft. high; the leaves closely resemble those, of S. rejlexum, but are slightly broader and less 
flaccid ; the open flowers measure a little less than an inch across, their pedicels being | to 1| in. long. The 
sepals are 3 lines long by 1 broad ; the short corolla-tube is | line long, the spreading lobes 5 lines long by 2 
broad. The truncate corona-segments are erect as in JS. reflexum and 2| lines long, overtopping the stalked 
gynostegium ; a membranous keel runs from the base of the teeth down the middle of the lobe. 
In flower in January. 

Distinguished from the Milanji species by its sturdier habit, larger flowers, rotate corolla, truncate coronar 
segments, and the shape of the arms of the translators. Is apparently near K. Schumann's ^S*. rJiacodes, which, 
however, has smaller flowers, much longer corona-segments, and the arms of the translators contracted with 
shoulders at the top. Deeaisne's Gomphocarpus chironioides, in DC. Prod. viii. p. 562, may belong to this species. 

Hah. Verj^ rare in elevated sandy thickets near Pedras do Guinga, Pungo Andongo ( Welwitsch^ ' Iter Angolense,' 
No. 4168). 

These plants belong to the genus Stathmostelma, recently established by Dr. K. Schumann (Engl. Bot. Jahrb. 
xvii. p. 129, t. 6). His diagnosis is based on the very broad arms of the pollen-carriers ("translators"), which 
characterize these species. Further investigation of the genera may lead, as Schumann himself points out, to a 
considerable reduction in their number, in which case Stathmostdma will probably be relegated to AsclepioB and our 
species will become respectively A. reflexa and A. Welwitschii, 



COLLECTED BY ME. ALEXANDER WHTTE. 29 

GoMPHOCARPUS PALUSTRis, K. Schum. in Engl. Bot. Jahrb. xvii. p. 127. 
Hah. Zomba. 

Margaeetta rosea, Oliver, in Trans. Linn. Soc. ser. I. xxix. p. 111. 
Mab. Milanji. Nos. 106, 117. / 

Pergtjlaria sangtjinolenta, Lindl. ex Sims, in Bot. Mag. t. 2532. 
Hah. Zomba. 

MosTUyEA Brunonis, F. Didrichs. in Vidensk. Med. Nat. For. Kjobenh. 1853, p. 87. 
Hah. Milanji. 

BUDDLEA, sp. 

Hah. Milanji. No. 113. 

Seb^a crassul^folia, Cham. & Scblecht. in Linnaea, i. 193 {fide Schinz). 
Hah. Milanji. 

Trichodesma physaloides, a. DC. in Prod. x. p. 173. 
Hah. Zomba. 

Hewittia bicolor, Wight & Am. in Madr. Journ. Sc. ser. I. v. (1837) 22. 
Hah. Milanji. 

IpoMCEA OBSCTJRA, Ker, in Bot. Reg. t. 239. 
Hab. Zomba. 

Ipomcea fulvicaulis, Boiss, ex Hallier in Engl. Bot. Jahrb. xviii. p. 128. 

Hah. Milanji. 

Our specimens have much shorter petioles than the type (Abyssinia, Schimper, Sect, i 
p. 240), but agree with it in other respects. 

Convolvulus malvaceus, Oliver, in Trans. Linn. Soc. ser. I. xxix. p. 117. 
Hah. Milanji. No. 108. 

Halleria elliptica, Thunb. in Nov. Act. Tips. vi. p. 39. 
Hah. Milanji. 

BUCHNERA, sp. 

Hah. Milanji, 6000 ft. No. 71. 

Ctcnium ADOJfENSE, E. Mcy. ex Benth. in Comp. Bot. Mag. i. p. 368. 
Hah. MUanji, 2000 ft. No. 186, 

Thunbergla. Kirkiana^ T. Anders, in Journ. linn. Soc. vii. p. 19. 
Hah. Milanji. 



30 THE PLANTS OF MILANJI, NYASA-LAND, 

Thijnbergia LANCiroLiA, T. Anders. 1. c. 
Sab. Zomba. 

Thtjnbergia alata, Eoj. ex Hook. Exot. PI. t. 177. 
Mob. Milanji. 

MiMiJLOPSis SESAMOEDES, S. Moore, sp. n. M. caule ascendente ? f olioso, foliis pro genere 
magnis, longipetiolatis, e basi sat attenuata, levissime cordata, ovatis acuminatis, 
grosse ac inapariter dentatis, membranaceis, inflorescentia e paniculis terminalibus 
brevibus (ac revera quam folia brevioribus) paucifloris vel elongatis et plurifloris 
composita ; paniculi ramis bibracteatis, bracteis lineari-lanceolatis calyce brevioribus, 
calycis laciniis anguste linearibus elongatis, laciniis anticis quam reliqua brevioribus, 
lacinia postica longiore ; corollse pro genere magnse, tubo superne gradatim ampli- 
ficato, limbi lobis rotundato-ovatis, lobo antico ma j ore, lobis reliquis subsequalibus ; 
staminibus inclusis, filamentis per paria in membranam latam pilosam fere ad basin 
tubi decurrentem connatis; antheris magnis, staminum anticorum loculo altero 
abrupte recurvo-calcarato, altero pauUo minore, mutico, staminum posticorum quam 
antica pauUo minorum loculis muticis ; disco obsolete ; ovario oblongo, basi subito 
incrassato ; stylo piloso coronato ; stigmatis lobo altero fere omnino obsoleto, altero 
satis elongato ; ovulis quovis in loculo 4 ; capsula anguste oblongo-ovoidea, calycem 
subaiquante, tetrasperma. 
B.ah. Milanji. No. 89. 

Caulis superne ultra 0*2 cm. diam., tetragonus, arete pubescens, mox glaber ; foliorum 
lamina usque ad 120 cm. longa (folia minora vero exstant), juxta medium 6'0-6"5 
cm. lata, ima basi fere ad I'O cm. angustata, utrinque, prsesertim in nervis, minute 
puberula ; folia subtus decoloria, juvenilia pubescentia ; petioli 3'0-vix 6'0 cm. longi, 
pubescentes, cito puberuli ; paniculae glanduloso-piloso-pubescentes ; calycis laciniie 
majores l*5-2'5 cm. loDgge, obtusae, nequaquam spathulatse, glanduloso-pilosse ; 
corolla 30 cm. diam., forsan violacea; limbi lobus anticus 1*1 cm. latus, lobi 
reliqui 0*8 cm. ; tubus 1*7 cm. longus, sub faucibus circa 1'5 cm. latus, deorsum ad 
05 cm. constrictus, antice pilis decurvis longiusculis obsitus ; filamenta libera 
complanata, staminum anticorum vix 06 cm. longa, staminum posticorum paullo 
breviora ; membrana decurrens 06 cm. longa ; antherarum apice breviter pilosarum 
et obtusaruni loculi majores 06 cm. longi, calcar eorum 01 cm. longum, ad angulum 
rectum loculo insidens ; ovarium 06 cm. longum, fere glabrum, apice obtusum ; 
stylus 1'3 cm. attingens ; stigmatis lobus alter 0*4 cm. longus, alter ad 002 cm. 
reductus ; ovula 01 cm. diam., biserialia ; capsula usque 2*5 cm. longa, breviter 
rostrata ; semina suborbicularia, 0*7 cm. longa, 0*8 cm. lata, brunnea, laevia, reti- 
naculis rectis sat validis 025 cm. longis fulta. 
Species memorabilis, et ob folia magna grosse dentata necnon tiores majuscules conge- 
neribus ab omnibus plane distincta. 



COLLECTED BY ME. ALEXANDER WHTTE. 31 

Crossandra Geeenstockii, S. Moore, in Journ. Bot. 1880, p. 37. 
Hab. Milanji, 6000 ft. No. 93. 

AsTSTASiA COROMANDELIANA, Nees, in WalL PI. As. Rav. iii. p. 89. 
Hah. Zomba. 

Brachtstephanus africanus, S. Moore, sp. n, B. caule folioso sat valido, in longitu- 
dinem striate, appresse pubescente, parumper glabro ; foliis majusculis, longius- vel 
brevipetiolatis, ovatis, elongate caudato-aeuminatis, basi in petiolum gradatim 
desinentibus, utrinque puberulis et in nervis pag. inf. appresse ac strigose pubescen- 
tibus ; spicis ramulos laterales breves terminantibus (an nnqnam verc axillaribus ?) 
elongatis, brevissime pedunculatis, vix omnino sessilibus ; bracteis majusculis, 
imbricatis, rliombco-ovatis, longe acuminatis, puberulis ; bracteolis linearibus 
acuminatis, carinatis, bracteis circa duplo et calyce triplo brevioribus ; calycis alte 
5-partiti segmentis insequalibus linearibus acutis; corolla* majusculoe tubo ad 
fauces parum ampliato, labio antico brevissime trifido, labio postico integro ; stami- 
nibus longe exsertis, basi liberis ; stylo exserto, stigmate capitulato coronate. 
Sab. Milanji, 6000 ft. No. 56. 

Caulis 0*4 cm. diam., ad nodos tumidus ; intemodia saltern superiora brcvia, raro ultra 
30 cm. longa (exstat vero aliud usque ad 6*0 cm.) ; foliorum lamina 12'0-li'O cm. 
longa, 5*5-6'0 cm. lata (occurrunt equidem folia minora — 3'0-60 cm. — majoribus 
mixta) ; petioli ad 4*0 cm. longi, fol. supremorura attamen vix 05 cm. attingunt ; 
fol. omnia tenuiter membranacea, in sicco virentia ; costae secundaria? utrovis 
latere circa 12, alternatim insertae, raro suboppositae, leviter arcuatae ; costaB omnes 
pag. inf. marginibus elevatis pubescentibus instructae ; spicae usque ad 22'0 cm. 
longae, circa 0*7 cm. latae, puberulae ; bracteae circa 1*0 cm. longae, 06-0'7 cm. latae, 
firmae, leviter striatae, demum subdistantes ; bracteolae 0*5 cm. long., puberulae; 
calycis segmenta duo antica reliqua pauUo excedentia et re vera fere 1-5 cm. longa 
(reliqua IS cm.), omnia trinervia, dorso pubescentia; corollas puberulae tubus vix 
3"0 cm. longus, 012 cm. latus, ipsis in faucibus ad 0*25 cm. amplificatus ; labium 
anticum late ovatum, 08 cm. longum, 0*5 cm. latum, dentibus brevissimis (scilicet nee 
ultra 0'05 cm. longis) triangulari-acutis, trifidum; labium posticum oblongo-obo- 
vatum, ambo margine leviter crispulo-undulata ; filamenta faucibus ipsis inserta, 
complanata, 30 cm. longa ; discus brevissime bilobus, 04 cm. altus ; ovarium ovoideo- 
oblongum, discum triplo excedens; stylus filiformis, staminibus subaequilongus ; 
capsulae non suppetebant. 
Cum B. Lyallii, Nees, congruens potius quam cum B. cuspidatOj Baker, cujus bracteae 
ampKssimae et stamina subinclusa. Ceterum B. Lyallii habitu humiliore, foliis multo 
minoribus, bracteis angustis, floribus minoribus gaudet. Ob staturam elatam et foUa 
ampla propius accedere videtur ad speciem nondescriptam Madagascariensem 



32 THE PLANTS OF MILANJI, NYASA-LAND, 

{Humblot, n. 632) in Herb. Kew. Lasserv., bracteis vero diversis et floris partibus 
minoribus. 

This is an interesting addition to a genus hitherto known only from Madagascar. 

JiJSTiciA (§ Rostellaria) Whytei, S. Moore, sp. n. Herba foliosa, strigosc pubes- 
cens, caule erecto, tetragono; foliis firmis, brevipetiolatis, lanceolatis vel lineari- 
lanceolatis obtusis, deorsum angustatis ; floribus mediocribus axillaribus brevissime 
pedunculatis, bracteolis calyce multo brevioribus ; calycis segmentis 5 subsequa- 
libus, lanceolatis acuminatis, dorso carinulatis, pubescentibus ; coroUae tubo calyccm 
duplo excedente, recto, lato, faucibus dilatatis, labio postico bifido quam anticum 
amplum breviore; antherarum loeulo inferiore superiorem ecalcaratum paullo 
excedente ; disco integro ; ovario anguste ovoideo-oblongo complanato, sursum in 
styliun inferne appresse pilosum desinente; loculis biovulatis; capsula oblongo- 
ovoidea, superne gradatim angustata, hirtulo-pubescente, tetrasperma. 
Sab, Milanji. No. 136. 

Caulis verisimiliter parce ramosus, vix usque 0*2 cm. diam. Eolia 2*0-4'0 cm. longa, 
0'5-l-2 cm. lata, margii^e revoluta, integra ; costae secundarisB 4-6, oblique insertas 
leviter arcuatae ; petioli 0'2-0-5 cm. longi; pedunculi circa O'l cm. longi; bracteas 
foliaceae, oblongSB obtusae, 0-7 cm. longae, 022 cm. latae; bracteolae 0*07 cm. longae,- 
calycis segmenta 0*25 cm. longa, 1-nervia; corollas extus pubescentis verisimi- 
liter dilute purpureas tubo 0*45 cm. longo, ima basi 0*1 cm., sub faucibus vero 
0*22 cm. lato ; labium posticum oblongum, erectum, 0'3 cm. longum ; labii antici, 
0*4 cm. longi et vix totidem lati, lobus intermedins late ovatus, lobi laterales oblongi, 
ille 0*15 cm., hi equidem 01 cm. lati ; palatus maxime cminens ; filamenta 
planata, basi leviter puberula, vix 0*3 cm. longa ; antherarum loculus inferior circa 
0*07 cm. longus, ejus calcar angulo recto insertum ; connectivus pilosus ; discus 
0'03 cm. altus; ovarium O'lo cm. longum, stylo obtuso 0-5 cm. longo coronatum; 
capsula 055 cm. longa, 023 cm. lata ; semina parva, oblongo-cylindracea, crebro 
tuberculata, pallida, 0*07 cm. longa, retinaculis debilibus oblongis obtusissimis, 
comparate latis, nervosis iisdem brevioribus suffulta. 
Mihi videtur J, neglectce, T. Anders., aiiinis, cujus attamen folia diversiformia, flores 

majores, calycis segmenta multo longiora, &c. 

JtrsTiciA Anselliana, T. Anders, in Journ. linn. Soc. vii. p. 44. 
'M.ab, Milanji. 

JusTiciA (§ Horniera) melampyrtjm, S. Moore, sp. n. Erecta, foliosa, caule tetragono, 
minute necnon appresse puberulo ; foliis membranaceis, lanceolatis vel ovato-lanceo- 
latis, obtuse acutis vel acuminatis, basi cuneatis, petiolis brevibus fultis, fere omnino 
glabris vel secus nervos pag. inferioris appresse pubescentibus ; floribus mediocribus 
axillaribus solitariis vel faucibus quavis ex axilla oriundis, brevissime peduncu- 
latis ; bracteis bracteolisque exiguis, calyce multo brevioribus ; calycis segmentis 



COLLECTED BY MK. ALEXANDEE WHYTE. 33 

lanceolatis, longe subtiliter acuminatis, ciliolatis ; corolla? tubo Icvitcr curvato, calyci 
subaequilongo, sat lato, faucibus vero admodum coarctatis ; limbi labio postico bifido 
quam anticum amplum breviter trilobum manifestc breviore ; antlierarum loculo 
superiore brevissime mucronulato, longiorc ac angustiore ; disco intcgro ; ovario 
oblongo, sursum gradatim angustato ; stylo obtuso, deorsum apprcsse piloso ; loculis 
biovulatis ; capsulis more sectionis dimorphis, aliis longioribus oblongis, sursum 
sensim coarctatis, aliis parvis oblongo-ovoideis, apice rostratis, tetraquetris, angulis 
cristulatis. 
Bab. Milanji. No. 135. 
Caulis 015-0-25 cm. diam., ejus anguli saltem in sicco eminentes ; foliorum lamina 
2*5-4*5 cm. longa, 0-8-2'3 cm. lata ; costae secundarioe 5-6, angulis angustis insertae, 
fere rectae ; petioli 0*7 cm. longi ; bracteae et bracteola? lincares, acutiuscula), 013 cm. 
longse ; calycis segmenta 0'4 cm. longa, dorso obtuse carinata ; corollae extus 
minute pubescentis tubo ima basi 0"18 cm. lato, faucibus vero 0*28 cm. lato; 
labium posticum ovatum 0*33 cm. longum ; labii antici, 0*4 cm. longi, lobi 0*1 cm. 
longi, lobus intermedins quam lobi laterales paullo latior ; lilamcnta crassiuscula, 
0*33 cm. longa ; antherarum loculus superior 008 cm. longus, inferior duplo longior, 
hujus calcar curvatum et apice bifurcatum; discus 0*05 cm. altus; ovarium 
0*15 cm. longum ; stylus 05 cm. longus; capsular majores 06 cm. longae, 0*2 cm. latae, 
acutae, verisimiliter 2-4-(an 2-3?)spermaD ; minores 043 cm. longae, media 02 cm. 
lata, 1-spermae, omnes obsolete puberulae ; semina ambitu ovata, brunnesccntia, 
crebro tuberculata, 0"1 cm. longa. 
A J. insula7'i, T. Aud., secernenda ob folia angustiora brevius petiolata, bracteas 
longe dispares, calycis segmenta longiora, cariuata, nee longe ciliata, flores aliquatenus 
majores et corollas lobos majores praebentes, antherarum loculum inferius elongatum 
et calcare bifurco nee siniplici auctum, ovarium longius et sursum angustatum, stigma 
obtusum nee leviter capitulatum, necnon capsulam breviorem. 

IsoGLOSsA MiLANJiENsis, S. Moorc, sp. n, Caule valido ascendente, geniculate, sur- 
sum folioso ; foliis amplis, longipetiolatis, late obovato-oblongis, caudato-acuminatis, 
prseter nervos pubescentes, obscure puberulis ; floribus mediocribus in spicis brevibus 
terminalibus vel axillaribus dispositis; bracteis obovatis acutis vel breviter vcl 
longius acuminatis, caly ce brevioribus vel cum subaequantibus ; bracteolis lanceolatis 
acutis, calyci subaequilongis ; calycis segmentis inaequalibus (anticis longioribus), 
lineari-lanceolatis acutis vel acutiusculis, ciliolatis ; corollae tubo ample cylindraceo, 
calyce breviore ; labii antici lobo medio obovato, lobis lateralibus oblongis breviore, 
labii postici elongati lobis brevibus, late oblongis ; staminibus subexsertis ; antheris 
discretis, distantibus, transversis ; ovario ovoideo-oblongo, stylo crassiusculo, stami- 
nibus aequilongo ; capsula baud visa. 
Sah, MHanji, 6000 ft. 

Caulis 0-4 cm. diam., ad nodos tumidos usque ad 0-8 cm. diam., obtuse angulatus, 

SECOND SERIES. — BOTANY, VOL. IT. ^ 



M THE PLANTS OF MILANJI, NYASA-LAND, 

appresse pubescens, deinde fere glaber ; foliorum lamina ad 20*0 cm. longa et 7*5 

cm. lata, temiiter membranacea, in nervis appresse et minute pubescens ; costae 

laterales utroque latere 10-12, superiores distantius insertse, levissime arcuatse ; 

petioli ad 6'5 cm. longi, complanati, demum glabri; spicse nuUse ultra S'O cm. 

longge, densifloree ; pedunculi suffulcientes ascendentes, 1"0 cm. longi, et una cum 

spicis pubescentes ; bractese circa 05 cm. longae, ciliolatse ; bracteolse ciliatae, 0*6 

cm. longae ; calycis segmenta antica 0'7 cm. longa, segmenta reliqua pauUo breviora 

et obtusiora, omnia puberula; alabastrum aperiens pauUo ultra 1*0 cm. longum; 

coroUse tubus vix 0*5 cm. longus et latus ; palatus convexus, venoso-rugosus ; limbus 

circa 1'3 cm. diam. ; labii antici lobus medius 0*45 cm. longus, lobi laterales 0-55 cm. 

longi, margine ad lobum medium spectante dente auctis vel ibidem expansis ; labii 

postici, 1*0 cm. longi, lobi 033 cm. longi ; lobi omnes obtusi ; filamenta juxta 

basin tubi inserta, 0-7 cm. longa, puberula ; connectivus O'l cm. longus ; antherae 

oblongse, 0'15 cm. longae, omnino ecalcaratse ; discus crenellatus, 0*05 cm. altus ; 

ovarium 0*17 cm. longum, superne in stylum pauUo incurvum 0'7 cm. longum desinens. 

Species distincta, nulli affinis. Isoglossa g7'andiflora {Ecteinanthus grandiflorus, 

T. And.) folia minora brevipetiolata, calyces granduloso-tomentosos, ilores multo 

majores ostendat, et I. laxa^ Oliv., insuper folia parva, flores laxe pauiculatos : forsan 

proxime accedat ad I. origanoidem {Rhytiglossam, origanoidem, Nees), cujus habitus et 

folia sat similia, sed species Neesiana spicas densiusfloras, bracteas omnino dispares, et 

flores minores prsebet. 

Htpoestes vERTiciLLATiis, R. Br. Prod. p. 474. 
Hab. Milanji, 6000 ft. No. 151. 

Htpoestes pha.ylopsoides, S. Moore, sp. n. Caule foliato tetragono pubescente, 
dein glabro ; f oliis ovato-lanceolatis acutis vel breviter acuminatis, margine repandis, 
supra appresse strigoso-hirtulis, subtus summum puberulis, petiolis sat longis insi- 
dentibus ; capituHs terminalibus vel axillaribus ambitu ovatis vel subhemisphsericis, 
plurifloris ; bracteis exterioribus foliaceis, amplis, late deltoideo-ovatis, interioribus 
oblanceolatis, omnibus obtusis, pubescenti-hirtulis ; bracteolis parvis lanceolatis, 
calyce plus quam duplo brevioribus ; calycis seg mentis subaequalibus, anguste lineari- 
lanceolatis acuminatis, pubescenti-pilosis, hyalinis ; corollse calycem J-superantis 
tubo superne ampliato, ibique pilosulo ; labio postico brevissime bidentato, antico 
3-lobo, lobo medio ovato, lobis lateralibus lanceolatis ; staminibus quam labia 
brevioribus ; disco lobulato ; stylo exserto, apice breviter ac subaequaliter bifido ; 
capsula calycem pauUo excedente, oblongo-lineari acuta, ejus loculis 1-2-spermis. 
Sab. Milanji. No. 126. 

Caulis 20-30 cm. diam., longitrorsum striatulus ; internodia circa 7'0-9*0 cm. longa, 
ramulorum vero juvenilium multo breviora ; folia pleraque 40-7'0 cm. longa et 
2'0-4*0 cm. lata, basi in petiolum angustata ; costae secundariae utrovis latere 6-8, 
angulis latis insertae, aperte arcuatae; petioli tenues, 1'0-15 cm. longi, pubescentes; 



COLLECTED BY MR. ALEXANDER -WIIYTE. 3(^ 

capitula vix usque 20 cm. longa et 3*0 cm. diam., exstant vera minora ; bractese 
exteriores circa 1*2 cm. longac et latae, interiores usque VS cm. longa?, 3 cm. lat«, 
omnes basi dorso leviter carinulatae ; bractcoloe 03 cm. longa?, acutatjc, dorso 
pilosae; calycis segmenta usque 07 cm. longa, 008 cm. lata, 1-nervia ; corolla 
1-0 cm. longa, ejus tubus 0*45 cm. longus, sub faucibus 0'23 cm. latus ; limbi extus 
pubescentis labia sub.Tequalia, inferius vero paullulum minus ct 0-55 cm. longum 
(superius 0'6 cm. longum), hoc oblongum, illius lobi a^quilongi et circa 0*1 cm. longi; 
filamenta complanata, 25 cm, longa ; antherae late oblonga?, obtussisimai, vix 
0-2 cm. longse; discus 002 cm. altus; ovarium ovoideum, acutatum, transparens, 
0'12 cm. longum; ovula loculis multo minora ; capsulal-2 cm. longa ; ejus valvular 
a latere compressse, puberulae, subito acutatae, pars asperma 0*5 cm. longa et dorso 
rugosa ; semina purpureo-nigra, minute tuberculata, vix 0*2 cm. diam., retinaculis 
oblique truncatis, apice decoloribus, 0*2 cm. longis, suflfulta. 
Speciem nulli affiniorem, ob bracteas amplas, eas Phaylopsidis alicujus mentientes, prirao 

obtutu cognoscere potes. Porsan juxta S. diclipteroidem^ Nees, plantiim ^fadagas- 

cariensem, intercalanda sit ; nobis attamen videtur potius cum specie in Uerb. Kew. 

nondum descripta a cl. Wood in Natal (no. 1068) et in Orange Free State a cl. Cooper 

(no. 3035) lecta componenda. 

Selago Whyteana, Rolfe, sp. n. Pruticulosa, ramosissima; rami pubcscentes; folia 
linearia, subacuta, asperula, 3-6 poll, longa ; spicae breves, terminalcs, numero- 
sissimse ; bractese lineari-oblongae, obtusaB, concavae, la;ves, | lin. longai ; calyx 
^ lin. longus, 5-fidus ; lobi insequales, lineari-oblongi, obtusi ; corolla 1 lin. longa, 
5-fida ; tubus brevis ; lobi oblongi, obtusi. 
Bab. Milanji, 6500 ft. No. 28. 

This species has much of the general habit and characters of S. Dregei, Rolfe (Joum. 
Linn. Soc. xx. p. 353), but the leaves are longer and more slender, and the flowers less 
than half the size of those of that species. 

Selago milanjiensis, Eolfe, sp. n. Pruticulosa ; rami puberuli ; folia linearia, sub- 
obtusa, 5-9 lin. longa ; spicse breves, terminalcs, numerosae ; bracteae lineari- 
oblongse, obtusge, concavse, laeves, |-1 lin. longae; calyx i lin. longus, 5-fidus; 
lobi insequales, lineari-oblongi, obtusi, ciliati ; corolla 1 lin. longa, 5-fida ; tubus 
brevis ; lobi rotundato-oblongi, obtusi. 
Hab. Milanji. 

Allied to the preceding, but a smaller, less robust plant, with much less pubescent 
branches, laxer inflorescence, and the calyx-lobes much less strongly ciliate. 

Lantana s/iLVi^roLiA, Jacq. Hort. Schoenb. iii. p. 18. 
JELab. Milanji. 

f2 



B6 THE PLANTS OF MILANJI, NYASA-LAND, 

ViTEX MiLANJiENSis, Britten, sp. n. Arbor ; ramis tetragonis, puberulis ; foliis 5-folio- 
latis ; foliolis oblongis truncatis mucronatis, basi cuneatis, margine integris, sinuatis, 
supra scaberulis, subtus pallidioribus, pubescentibus, nervo medio venisque primariis 
conspicuis ; lateralibus breviter, central! longiuscule petiolulatis ; pedunculis axil- 
laribus folio brevioribus, apice dichotome cymosis ; bracteis subulatis ; calyce 
hirsuto, 5-lobato, lobis deltoideis acutis ; staminibus inclusis. 
Bab. Milanji, 6000 ft., No. 138 ; Zomba. 

In general appearance this much resembles V. madiensis, Oliver, of which it may 
possibly be a variety, especially in the inflorescence. It differs in the number and shape 
of the leaflets, which are quite entire, and is less hairy in all its parts. 

Clerodendron discolor, Vatke, in Flora, xliii. p. 536. 
Bab. Milanji, No. 124 ; Zomba. 

Clerodendron myricoides, E/. Br. 
Bab. Milanji. Nos. 126, 193. 

Plectranthtjs sanguineus, Britten, sp. n. Caule erecto, rugoso, inferne glabrato, supeme 
farinoso, ramoso ; foliis post anthesin evolutis (speciminibus immaturis) ; verticillas- 
tribus densis, multifloris ; floribus subsessilibus, calyce exteriore villoso, cam- 
panulato, 1^ lin. longo, dentibus acuminatis, setibus luteis articulatis ; corolla 5 lin. 
longa ; seminibus minutis, glabris. 
Bab. Milanji, 6000 ft. No. 46. 

A very distinct species, allied to P. cylindraceus, Hochst. Stems upright, 2 feet 
high, softly woody, giving off small lateral floral branches, and terminating in a panicle 
of 4-7 racemes 3-6 in. long ; the central are longest. Verticillasters of the upper branches 
densely crowded together, although not so much so as in P. cylindraceus ; those of the 
lower branches more distant. Calyx campanulate, with acuminate teeth, shaggy without, 
with yellow jointed hairs, and exuding, like the bud -scales, corollas, and stamens, bright 
red drops apparently of a resinous nature. Leafy shoots developing later from the lower 
buds of the main stem ; leaves immature in specimens seen, tomentose, rugose, crenate, 
roundish or wedge-shaped, J— | in. long, sessile or very shortly petiolate. 

Plectranthus elegans, Britten, sp. n. Caule erecto ; foliis petiolatis, ovatis, crenatis 
acutis, basi cuneatis ; panicula racemosa, elongata, ramosa ; cymis oppositis breviter 
pedunculatis, 2-3-floris ; calyce dentibus insequalibus, superiore obtuso ovato, 
cseteris acuminatis; corollse labio superiore trilobo, lobo centrali truncato, lobis 
laterahbus rotundatis, labio inferior e concave. 
Bah. Milanji. 

A perennial erect herb (upper portion alone seen) ; stem furrowed, hairy ; leaves 
shortly petiolate ovate, 1-2 in. long, 1 in. or more broad, tapering to a wedge-shaped 
base, broadly and bluntly crenate, acute, scabrous and dark green above, softly hairy 
and paler beneath, with prominent veins ; panicle narrowly triangular, leafless, racemose, 



COLLECTED BY ]\IE. ALEXANDER WHTTE. 37 

slender, 7 iDches or more long ; bracts small, entire, ovate acute ; branches 3-4 in. long, 
ascending ; cymes nearly sessile, 2-3-flowered, internodes | in. or more long ; pedicels 
2 lin. long ; calyx minutely hairy, broadly campanulate, 1 lin. long ; upper tooth ovate 
obtuse, lower teeth acuminate, the two lowest the longest ; corolla 3-4 lin. long, upper 
lip 3-lobed, centre lobe truncate, 2 lateral lobes rounded, lower lip concave ; stamens 
exserted. 

Plectranthus Melleri, Baker, in Journ. Bot. 1882, p. 243. 
Hah. Milanji, 6000 ft. No. 102. 

Plectranthus mangantensis. Baker, in Oliver, Fl. Trop. Afr. iv. ined. 
JSah. Zomba. 

Plectranthus glandulosus. Hook. fil. in Journ. Linn. Soc. vi. p. 17. 
JSab. Zomba. 

OciMUM eilamentosum, Forsk. PI. ^gypt. Arab. p. 108, var. ? 
Mah. Milanji. No. 178. 

OciMUM HiANS, Benth. in DC. Prod. xii. p. 36 ? (0. caltcosum, Hochst. in SchiiTip. 
PI. Abyss, no. 2303). 
Hab. Milanji. 

Orthosiphon coloratus, Vatke, in Linnsea, xliii. p. 86. 
Hab. Zomba. No. 138. 

Orthosiphon Kirkii, Baker, in Fl. Trop. Afr. iv. ined. 
Hab. Milanji. No. 101. 

Hoslundia decumbens, Benth. in DC. Prod. xii. p. 54. 
Hab. Milanji, 6000 ft.. No. 156 ; Zomba. 

Htptis pectinata, Poit. in Ann. Mus. Par. vii. (1806) p. 474, t. 30. 
Hab. Zomba. 

MiCROMERiA BiFLORA, Benth. Lab. p. 378. 
Hab. Milanji. No. 84. 

Scutellaria Livingstonei, Baker, in PI. Trop. Afr. iv. ined. 
Hab. Zomba. 

Stachys ^thiopica, Linn. Mant. p. 82. 
Hab. Milanji. Nos. 20, 128. 
A wide-spread South- African species, not hitherto, I believe, recorded from the 

Tropics. 

Leucas glabrata, B. Br. Prod. p. 504. 
Hah. Milanji. Nos. 127, 128, 191. 



38 THE PLANTS OF MILAN JI, NYASA-LAND, 



APETAL^. 
PiPEEACE^ and ErPHORBiACE^. (By E. G. Baker, F.L.S.) 

AmARANTACE^, POLYGONACEiE, PrOTEACEJ]:, ThYMELEACE^, SANTALACE^y 

and Mtricace^. (By A. B. Rendle, M.A., E.L.S.) 

Peperomia reflexa, Dietr. ; DC. Prod. xvi. 1, p. 451. 
Sah. Milanji. 
Widely spread in the Tropics. 

Phtllanthus rotundifolius, Willd. ; Miill. Arg. in DC. Prod. xv. 2, p. 405. 
SLah. Milanji. 
Widely spread in the Tropics of the Old World. 

Phyllanthtjs rotijndipolius, var. leucocalyx, Miill. Arg. 1. c. p. 406. 
Hab, Milanji, alt. 6000 ft. No. 23. 
Distr. Abyssinia ; Zambesi-land ; Angola. 

Phyllanthtjs, sp. 

Sab. Milanji, alt. 7000 ft. No. 29. 
There are no flowers on this specimen. 

Antidesma, sp. 

Bab. Milanji. No. 170. 

AcALYPHA viLLicAULis, A. Rich. Tent. El. Abyss, ii. p. 248 ; Miill. Arg. 1. c. p. 845. 
Hab. Zomba ; Milanji, alt. 6000 ft. No. 120. 
Distr. Abyssinia, Central Africa, Manganya Hills. 

AcALYPHA bengtjelensis, Miill. Arg. in Jonrn. Bot. 1864, p. 335. 
Mab. Zomba ; Milanji, alt. 6000 ft. No. 65. 
Previously collected only by Welwitsch in Benguela. 

Euphorbia shirensis, Baker fil., n. sp. Radix lignosa ; caulibus herbaceis, erectis, 
striatis, glabris, simplicibus vel parce ramosis, sparse foliosis, rarissime subdense 
f Gliosis ; f oliis integris, obovatis vel anguste ovatis, vel oblongis alternatis subcoriaceis 
penninervatis, utrinque glabris, petiolatis, petiolis brevissimis, foliis ramorum late- 
ralium angustioribus ; foliis floralibus liberis ovatis vel rotundatis, latioribus quam 
longis, glandulis semilunatis ; filamentis glabris, sty lis erectis, stigmatibus bifurcatis ; 
capsulis pedunculatis, elevato-rugosis, trisulcatis ; semiuibus rubro-fulvis Isevibus. 
Mab, Milanji. 



COLLECTED BY MB. ALEXANDER WHTTE. 39 

Stems 9-15 in. high ; leaves |-1J or even 2 in. long ; flowering leaves J in. long. 
This plant belongs to the section Tithymalus, and is near E. dilatata, Hochst. 

Euphorbia Whtteana, Baker fil., n. sp. Fruticosa, ramis elongatis, virgatis, subdense 
foliosis, tenuiter striatis ; umbellae radiis 3-9 brevibus pluries bifidis ; foliis sub- 
coriaceis, integris, anguste oblanceolatis, superne glabris vel glabriusculis, subtus parce 
glaucescentibus, penninervatis, margine subrevoluto, apice in mucronem abruptiuscule 
brevissime attenuatis ; folii§ floralibus liberis, triangulari-acutis ; involucri turbinati 
glandulis crescentiformibus breviter cornutis; stylis bifurcatis ; capsulis pcdunculatis, 
glabris tenuiter rugulosis ; pedunculis pubescentibus ; seminibus non visis. 
Bab. Milanji plateau, alt. 6000 ft. No. 24. 
Cauline leaves about 1 in. long. 

This plant belongs to the section Tifhymalus^ and is allied to E. itwolucraia, E. Mey. ; 
the bracts, however, are much narrower than in this species, and the cauline leaves are 
nearly twice as long and rarely deflexed. 

Celosia trigyna, Linn. 
Sab. Milanji, 2000 ft., Oct. No. 166. 
Distr. Throughout Tropical Africa ; also Arabia and Natal. 

Polygonum serrulatum, Lag. Gen. et Sp. Plant. 14, no. 181. 

Hab. Zomba, Sept. 

Distr. Throughout the Mediterranean region ; Syria to North India ; North, Tropical, 
and South Africa, Canary Is., and Madeira. 

Protea ntas^, Hendle, sp. n. Caulescens ; foliis oblongis, apice rotundatis emarginatis, 

basi attenuatis, sessilibus, unicostatis, glabris, coriaceis ; capitulo terminali, magno, 

turbinato, squamis multiseriatis, arete imbricatis, laevibus, inferioribus minoribus 

late ovatis, margine ciliolatis, superioribus majoribus oblongis, deinde spathulato- 

oblongis, margine glabris ; perianthio glabrato vel sparsissime pubescente. 

Shoot strong and woody. Leaves (of last season) oblong, with rounded emarginate 

apex, narrowed at the base, sessile, one-nerved, glabrous, coriaceous ; 2-2} in. long, 

6-7 lines broad. Capitulum terminal on shoot of present season, from whicli all the 

leaves have dropped, turbinate, 3^ in. broad at the top, 2| in. long; bracts of involucre 

many-seriate, closely imbricate, smooth, the lower and smaller broadly ovate, with a 

ciliolate margin, the size increasing upwards, and the shape varying through ovate to 

the uppermost, which are spathulate-oblong ; except in the lower the margin is glabrous. 

Perianth 1} in. long, glabrous or very sparsely pubescent ; the lower broad portion of 

the tube 5 lines long by 1 line in diameter, the slender portion } in. by i line, the limb 

7 lines long, segments J line broad, linear with a subulate apex produced IJ lines 

beyond the linear anther, which is 4 J lines long. Style subulate. If lines long, channelled 

above. There is a hypogynous tuft of ferruginous hairs 5 lines long. 



40 THE PLANTS OF MILANJI, NTASA-LAND, 

Sah. Milanji. No. 30. 

Nearly allied to Frotea madiensis, Oliver (Trans. Linn. Soc. ser. I. xxix. p. 143), the 
glabrous capitulum being about the same size in each and having a similar arrangement of 
bracts ; but the leaves differ very markedly, those of F. madiensis being larger, oval, and 
more pointed, 7 in. long by 2^-2f broad, while the perianth-tube is densely villose. 
They are both allied to P. ahyssinica, Hochst., which, however, is at once distinguished 
by the white sericeous bracts, hairy perianth, and differently shaped leaves. 

Arthrosolen plavus, Uendle, sp. n. Frutex ramis sesquipedalibus, teretibus, striatis, 
glabratis, viridibus ; foliis sparsis, adscendentibus, inferioribus lineari-lanceolatis, basi 
longe attenuatis, superioribus lineari-subulatis, uninerviis, glabris, rugulosis ; invo- 
lucris squamis, multiseriatis, imbricatis, quam flores minoribus, glabris, exterioribus 
lanceolatis acuminatis, interioribus ovatis acutis vel subacuminatis ; capitulo semi- 
globoso, multifloro; perianthio flavo, tubo gracili minute pubescente, limbi lobis 
patentibus obovatis. 
Elowering branches of the present year only, IJ ft. long, terete, striate, glabrous, green. 
Leaves scattered, ascending, linear-lanceolate, below with a long attenuate base, 
becoming linear- subulate above, apex subacute, one-nerved, glabrous, rugulose, 8-10 lines 
long, J-l line broad. Involucral bracts multiseriate, imbricate, outer lanceolate, 
acuminate, inner ovate, acute or subacuminate, shorter than the flowers, glabrous, 
green, except the innermost, which are yellowish. Capitula hemispherical, many- 
flowered, J in. long by ^ in. broad. Pedicels ^-f line long ; a tuft, 1 line long, of stiff 
colourless persistent hairs surrounds the base of the perianth. Perianth yellow ; tube 
minutely pubescent, 4J lines long, J line ia diameter ; limb spreading ; lobes obovate, 
apiculate, minutely pubescent below, minutely papillose above, 2 lines long, more than 

1 line broad at the base. No scales on the perianth. Stamens sessile, in two whorls 
of four; anthers of upper row linear-oblong, acute, slightly exserted, f line long, of 
lower row oblong, with a rounder apex. Style short, filiform, about 1 line long. 
Stigma capitate, hairy. 

Sab. Milanji, 6000 ft. {A. Whyte, no. 99; Herb. Mus. Brit); Blantyre {L. Scott ^ 
Herb. Kew). 

Near A. glaucescens, Oliver (Journ. Lian. Soc. xv. p. 96), but the latter is a very 
glaucous plant, smaller, and has red flowers ; the perianth-limb, moreover, is lanceolate, 
IJ lines long, and less than ^ line at the broadest part, while that of A.Jlavus is obovate, 

2 lines long, and more than 1 line broad. 

Gnidia microcephala, Meisn. in DC. Prodr. xiv. p. 589. 

Hab. Milanji and Zomba, Sept. and Oct. 1891 {A. Whyte) ; Zomba and east end of 
Lake Shirwa, Oct. 1861, Livingstone's Zambesi Expedition (C. T. Meller; Herb. 
Kew) ; Namuli, Makua Country {J. T. Last ; Herb. Kew) ; Biggarsberg, 4000 ft. 
{J. M. Wood ; Natal Govt. Herb. no. 846) ; Macalisberg, Transvaal {Burke, no. 96) ; 
Kitui in Ukamba {Hildebrandt, no. 2807). 



COLLECTED BY MR. ALEXANDER WHYTE. il 

There is a very similar plant in Herb. Kew, collected on the Niger, W. Trop. Africa, 
by Dr. Baikie. 

Gnidia eastigiata, Rendle, sp. n. Suffrutex, subpedalis, ramis subfastigiatis, pubescen- 
tibus ; foliis densis, suberectis, rigidis, lanceolatis, acutis, imincrviis, glabris, supra 
rugulosis ; floribus solitariis, binis, aut forte ternis, in ramulis brevibus axillaribus 
sub vertice ramorum ; bracteis pubescentibus ; perianthio dense ciliato-pubescente, 
tubo bilineo cum pilorum fasciculo pone antheras superas, lobissubpatentibusovatis, 
squamis binis ligulatis, glabris, basi fultis; androBcio tetramero obdiplostamineo ; 
ovario ovato ; stylo lateral i filiformi ; stigraate capitate, saepius antheras inferiores, 
rarius superiores attinente, vel rare stylo brevissimo. 
A small shrub, up to 10 in. high, with ascending subftistigiate brandies, closely 
covered by the narrow, acute leaves. Leaves 5-6 lines long by J-1 line broad, the lower 
being the broader. Elowers on short, small-leaved axillary shoots, in the upper one or 
two inches of the branches ; the lowest shoots very reduced, with one terminal flower, 
the greater number more developed, with two flowers or the rudiment of a third, 
which in the larger ones may be fully developed ; at the top of the brancli again one- 
flowered. The small leaves of the shoots and the slightly larger bract are similar in 
form to the ordinary foliage-leaves, but arc pubescent like the perianth. Perianth- 
tube 2 lines long, J line broad, the slightly spreading limbs J line long, each with 
a pair of ligulate scales one each side of the base. Surrounding the base of the 
tube is a calycine circlet of stiffish hairs about J line long. A tuft of hairs behind the 
anthers of the upper row alternates with the scales. The lower persistent portion of 
the perianth-tube is of a darker colour and longitudinally ribbed ; the ovary in the flower 
reaches about halfway up this lower portion. The filiform style is attached a little 
below the apex, and varies in length in different flowers. 
Mab. Milanji, 6000-8000 ft., Oct. A very distinct species. 

Thesium ni&ricans, Eendle, sp. n. Suff'rutex, ramis et ramulis tenuibus, subrigidis, 
patentibus ; foliis sessilibus, linearibus, mucronatis, piano-con vexis, glabris, margine 
scabridulis; floribus parvis, apice ramulorum axillaribus, breviter pedicellatis ; 
bracteis cum duabus bracteolis foliis consimiiibus, bracteolis autem minoribus et 
florem fructumve subiequantibus ; perianthio campanuliformi, tubo brevi, lobis 4-5 
subpatentibus, glabris, apice inflexis papillosis, cum pilis paucis erectis pone anther- 
arum filamenta brevia ; stigmate capitate, stamina superante ; nuce ovoidea, rugis 
8-10 notata, cum reticulis paucis obliquis, quam perianthium latiore. 
A much-branched shrubby or suffrutesceut plant, which turns black in drying. The 
Hnear leaves flattened above, rounded beneath, from ^-f in. long, about i line wide, 
with the insertion continued as a ridge down the stem to the next leaf vertically below. 
There are four to five small flowers (1 Hne long) at the top of the branches, each in the 
axH of a leaf -like bract, with two shorter similar bracteoles near the base of the pedicel, 
at right angles to the plane of insertion of the bract, which is decurrrent to the next 

SECOND SERIES. — BOTANY, VOL. IV. ^ 



42 THE PLAJSTS OF MILANJI, NYASA-LAND, 

lower node; the base of the pedicel is adnate to the bract, and the bracteoles are 
inserted at the point where the two separate. The perianth-lobes are triangular- 
ovate, a little over ^ line long, and slightly involuted at the tip, more so in the fruit. 
The 4 or 5 stamens have very short filaments, behind which is a tuft of a few rather 
long upwardly -directed hairs. The ovoid nut is 8-10-ribbed, with a few oblique 
connecting reticulations, is 1 line long by f line in diameter, and broader than the 
persistent perianth. 

Hob. Milanji, 6000 ft., Oct. (no. 42); Zomba, Sept. 

Belongs to the section Euthesium, and is near Tkesium Wightianum, Wall., an East 
Indian species, from which it is easily distinguished by its subrigid habit, longer linear 
leaves, and smaller fruit. The flower of Th. Wightianum is half as large again, and the 
perianth-lobes are markedly spreading and papillose-pilose. 

Thesium (§ Frisea) Whyteanum, E/cndle, sp. n. (Plate IX. figs. 1-5.) Suffrutex, 

glaber, ramis tenuibus suberectis, pedalibus, infra nudis, supra foliosis, sub vertice 

ramulosis, ramulis foliosis vel floriferis ; foliis sessilibus, ramorum uncialibus, 

linearibus, crassiusculis, subteretibus, pungentibus, ramulorum brevioribus, plano- 

convexis ; floribus solitariis, axillaribus, vel singulis binisve in ramulis parvis ; 

bracteis et bracteolis foliis consimilibus, his autem minoribus, flore et fructu duplo 

brevioribus ; perianthii lobis quinque, apice involutis, cucullatis, et longe pilosis ; 

antheris glabris, pendentibus ; stylo brevissimo, stigmate papilloso ; nuce sub- 

globoso, decemcostato, oblique reticulato. 

A glabrous shrub with branches a foot long, leafless below, leafy above, branching 

near the top ; leaves of the primary branches subterete, longitudinally channelled, an 

inch or more long, of the secondary shorter, linear-subulate, flattened above, rounded 

beneath. Inflorescence apparently a long leafy spike ; the few uppermost flowers 

solitary in the axils of the leaves, those next below terminating short axillary shoots, the 

lowest in pairs, terminal and lateral on the shoots ; the bract and pair of bracteoles 

resembling the smaller leaves, the former about equal to the length of the flower, the 

latter about half. Plowers about 1 J lines long. Perianth cup and lobe nearly 1 line, the 

five lobes involute and hooded at the apex with a tuft of hairs, hairs continued about 

halfway down the margins ; the five glabrous anthers opposite the perianth-lobes fixed 

by a short filament and hanging down into the cavity ; a tuft of hairs springs from the 

perianth behind the filament and ascending is attached to the top of the anther. Style 

very short, with a papillose, scarcely swollen stigma, on a level with the base of the 

anthers. Eruit almost sessile, ovoid-globose or nearly globular, a little more than a line 

long and nearly as broad, crowned by the much narrower sub cylindrical perianth, with 

ten ribs, the alternate ones less pronounced, connected by irregular reticulations. 

Mab. Milanji, 6000 ft., Oct. No. 45. 

The discovery of this species extends the distribution of the well-marked section Frisea 
from South to Tropical Africa. 



COLLECTED BY MR. ALEXANDER WJIYTE. 43- 

Myrica PiLULiFEiiA, Eendle, sp. n. Arborescens ; foliis breviter petiolatis, ellipticis, 
acutis, serratis integrisve, uninerviis, supra laevibus, infra minute farinosis ; fructibus 
globosis in spicis axillaribus, sesquilineis ; pericarpio papilloso ; eudocarpio duro. 

Shoots of previous years woody, strongly marked with ridges decurrent from the 
prominent leaf-scar, and with conspicuous lenticels ; that of tlie present year farinose. 
Leaves shortly petiolate, elliptic, acute, serrate, one-nerved, glabrous and smooth 
above except for a few scattered hairs on the midrib, minutely farinose beneath ; the 
lower leaves of a shoot are much smaller, with an entire margin ; the larger leaves are 
from 2-2 J in. long, by 8-10 lines broad. Pruit on numerous spikes seated in the axils 
of the insertion of last year's leaves, all of which have fallen, globular, 1^ lines in 
diameter ; pericarp of papilla? with a waxy excretion ; endocarp hard, globuhir. 

Near Myrica salicifolia, Hochst., but differs in the regular shape and serration of the 
normal leaves, the smaller fruit, and the spherical endocarp, that of M. salicifolia being 
subovate or pyriform. The lenticels are also very large and characteristic. 

Sab. Milan ji, Oct. 

Pruiting specimens only. 

MONOCOTYLEDONS. (By A. B. Rendli:, M.A., P.L.S.) 

Cyperace^. (By C. Baron Clarke, M.A., F.ll.S.) 

EuLOPHiA LONGISEPALA, Rendle, sp. n. Herba caule tenui, sesquipedali, laxe paucifloro ; 

bracteis brunneis, brevibus, ovatis, acutis, pedicello perbrevi multo longioribus; 

sepalis longis, fere uncialibus, clavatis, apiculatis, petala longe excedcntibus ; 

petalis cuneato-spathulatis, apiculatis ; labcllo trilobo ; lobis lateralibus majoribus 

ad columnam ascendentibus, rotundatis, integris, venis prominentibus ; lobo terminah 

oblato, emarginato, margine crispo, verrucis lamelliformibus in venis mediis suflPulto ; 

calcare robusto, brevi, recurvato ; columna clavata, erecta, supra alata. 

The slender peduncle, IJ ft. long, bears a few almost sessile flowers arranged laxly 

in the axils of short bracts 2 lines long. The narrow clavate sepals are 11 lines long, 

nearly twice the length of the broader more spathulate petals, which are 6 lines long by 

If lines broad. The lateral lobes of the lip are 5 lines long, with thickened projecting 

lateral veins ; the shortened terminal lobe is 2^ lines long by 5 broad, bearing a double 

row of projecting lamelliform calli on the central veins. The short recurved spur is 

about J in. long. The erect column, broadly winged above, is 3 lines long, i line broad 

at the base, 1 line above. 

Near JE. Fetersii, Reichb. f., but at once distinguished by its slender clavate sepals much 
exceeding the petals ; E. Fetersii has also a broader, more ligulate petal, and a much 
narrower lip. Apparently, from the description, allied to E. Meleagris, Reichb. f., which 
has very unequal sepals and petals, but a many-(l7-18-)flowered raceme, much shorter 
oblong-lanceolate sepals, shorter, ovate-acute, irregularly denticulate petals, and a 
differently shaped lip. 

Mab. Milanji, Sept. 

g2 



44 THE PLANTS OF MILANJI, NTASA-LAND, 

EuLOPHiA MILAN JIANA, E/Giidle, sp. n. Caule valido, striate, foliis longis, lanceolatis, 
basi circumdato ; bracteis sterilibus membranaceis, pallide brunneis, margine longe 
connatis, internodia infera omnino, supera auteni partim, vaginantibus ; racemo 
dense corymboso, bipollicari, bracteis florentibus longis subulatis, pallide brunneis, 
venis fuscis striatis, ovarium pedicellatum excedentibus ; sepalis petalis consimilibus, 
sed majoribus, ovatis acutis, apice subfalcatis ; labello trilobo, lobo medio quam 
lobi laterales, oblongi, ascendentes, integri, glabri, plus triplo majore, obovato- 
orbiculari ; disco tomentoso, basi verrucis geminis suffulto ; calcare brevi obtuso ; 
columna brevi crassa, alls tenuibus. 
The peduncle springs from among long, lanceolate, longitudinally-ribbed radical leaves, 
with a long, narrow, sheathing base, immature in the specimen, but apparently resembling 
those of Eulojphia {^Orthochilus) ahyssinica^ Reichb. f . Only the upper part of the mature 
peduncles was present, 1 foot long by \ in. diameter below, and striate, with long, acute, 
sheathing bracts, which are membranous, pale brown, with fine veins, and their edges 
connate for more than half their length, slightly spreading at the tip, the two lower 
3-3J in. long, the uppermost 25-2J in. The inflorescence forms a dense corymb, 
2 in. high by 2 in. broad near the base. The flowering bracts are subulate, pale 
brown, marked with numerous darker veins, and exceed the stalked ovary ; the lower- 
most 14 lines long, the size decreasing upwards. The outer flowers are the larger. The 
sepals are 10-13 lines long by 4-6 broad ; the petals are smaller, 8 by 3 lines, but other- 
wise similar. The small oblong lateral lobes of the lip, 3 lines long by 2 broad, are 
entire and glabrous ; the central, almost suborbicular, lobe measures 6 by 5J lines, has a 
long-haired tomentum on the disk, with two knob-like warts on the veins just at its union 
witli the lateral lobes. The short thick column is 2 lines long by 1 wide, and has thin 
wings ; the short blunt spur is 2 lines long. 

Hab. Lower Milanji, 2,500 ft., Oct. 1891 (A. Whyte, no. 257) ; Manganja Hills 
(Livingstone's Zambesi Exped. 1861 ; Herb. Kew.) ; Livingstone's S. African Exped. 
(Herb. Kew.) ; Uganda, 1880 {Hev. C. T. Wilsoti, no. 114 ; Herb. Kew.) ; Nyasa-land, 
1891 {Buchanan, nos. 349 & 1328; Herb. Kew.). 

Near E. bicolor, Reichb. f., but differs in its slightly smaller flowers and corymbose 
inflorescence, that of B. hicolor being a somewhat crowded raceme, twice the length ; 
also in the lip, which is more delicate and devoid of the purple colour on the lateral lobes 
which characterizes those of E. bicolor, where moreover the long-haired tomentum is 
absent, scattered tufts of short filiform calli taking its place. 

It is also near -£'.(§ OrtJiochilus) abyssinica, Keichb. f., but distinguished^by the shape of 
the lip as well as by its larger flower and more crowded inflorescence. 

EuLOPHiA NvASiE, Ucndle, sp. n. Herba pedunculo sesquipedali, basi folio vaginante cir- 
cumdato, infra tereti, supra sulcato, bractea triangulari-ovata acuta, supra basin 
breviter connatam libera, pedunculum autem arete amplectante ; racemo laxo secundo 
9-pollicari, bracteis subulatis, patentibus, uninerviis, pedicella brevia paullo exceden- 



COLLECTED BY MR. ALEXANDER WHTTE. 45 

tibus ; sepala quam petala longioribus, oblongo-ligulatis, subacutis, petiilis ovalibus, 
obtusis ; labello semipollicari, parte inferiore et majore trapezoideo, superiore tri- 
angulari-ovato obtuso, dense tomentoso ; calcare brevi obtuso ; columna crassiuscula. 
Peduncle 19-21 in. long, clothed at the base with a sheathini^-leaf 2h in. Ion-', which 
also includes two young foliage-leaves ; 1 J line in thickness an incli above the base, 
terete below, becoming longitudinally grooved above when dry, bearing one bract at 
about a third the way up. Bract | in. long, margins connate for 1 line at the base, 
then rapidly receding to form the upper free triangular part, which, however, is closely 
applied to the peduncle. The lax secund raceme is about in. long ; the lower bracts 
are 5 lines long, the size decreasing upwards. The short pedicels are about 2 lines long. 
The oblong ligulate sepals are longer and much darker than the petals, 3-norved, 8 lines 
long, IJ broad, but narrower at the base ; petals 5 J by 2 lines, with a prominent midrib 
and two upwardly branching lateral veins. The lower part of the lip, when flattened, 
is trapezoid (the sides naturally ascend), 4 lines long by 2J wide in the lower half, then 
slightly widening upwards to 3 lines at the two rounded angles ; the apical lobe is 2 lines 
long by If broad at the base, and has its upper surface covered, except ncjir the margin, 
with a matted tomentum arising from the veins and passing down the three mc^dian nerves 
a short way into the lower part. The spur is blunt, swollen and rounded at the tip, 
2J lines long. The column is 2f lines long, 1 broad. 
Sab. Milanji, Sept. 

The structure of the flower recalls that of a group of Cape species including Eulophia 
lamellata, Lindl., which it somewhat resembles, but is distinguished by the marked 
difference in size between the sepals and petals, and the shape of the apical lobe of 
the lip, which in E. lamellata is much larger, quadrate, and has a longer tomentum ; 
the inflorescence in that species is branched and the bracts are longer. 

LissocHiLUS ARENARius, Lindl. in Journ. Linn. Soc. vi. p. 133. 

Lindley described this species from a West Tropical African plant (Barter, no. 1488), 
but it seems common in the East. 

Hab. Im Lande der Djur (Schweinfurth, Reise nach Central Africa, no. 1864) ; Nyika 
Country (Hev. K T. Wakefield ; Herb. Kew.) ; Lake Tanjanika {Lieut. Cameron ; Herb. 
Kew.); Habai Hills, Mombaz, March 1886 (Bev. W. E. Taylor-, Herb. Mus. Brit.); 
Milanji, Nyasa-land, Oct. 1891 [A. Whyte). 

Reichenbach (Otia Bot. Hamb. p. 75) makes Hildebrandt's plant, no. 1950, Mombassa, 
1876, the same species ; it difi'ers slightly from the other East Tropical specimens seen 
in its acuminate lateral sepals, and in having small warts on the surface of the labellum 
in front of the pairs of calli. The lamelliform calli vary in the different specimens 
in shape, in the character of the apex, which may be crenate or emarginate, and in their 
more or less erect position on the veins. 

LissocHiLUs LiviNGSTONiANUS, Rcichb. f. Otia Bot. Hamb. p. 114. 

Sab. Manganja HiUs, Livingstone's Zambesi Exped. {C.Meller; Herb. Kew.) ; Man- 



4i6 THE PLANTS OF MILANJI, NYASA-LAND, 

gaiija Hills, E. Africa (II. Waller; Herb. Kew.); Nyika Country, E. Trop. Africa 
[Rev. E. T. Wakefield', Herb. Kew.}; MHanji, Nyasa-land, Oct. 1891 (A. Whyte). 
An East Tropical African species. 

LissocHiLTJS Wakefieldi, Reichb. f. & S. Moore, in Journ. Bot. 1878, p. 136. 

Hab. Mombassa, Kildehrandt \ Rabai Hills, Mombaz {Hev. W. E. Taylor; Herb. 
Mus. Brit.); Milanji, 1891 {A. Whyte). 

LissocHiLUS MiLANJiANUS, Bcndle, sp. n. Pedunculo elongato, racemoso, laxifloro; 

bracteis membranaceis, brunneis, triangulis, longe acuminatis, ovaria pedicellata 

baud sequantibus ; sepalis lateralibus ad basin labeili adnatis, oblongis, apiculatis, basi 

truncatis ; dorsali oblongo angustiore ; petalis orbicularibus, minute apiculatis, multi- 

nerviis ; labeili lobis lateralibus brevibus, truncatis, latera sacci cum calcare con- 

tinui formantibus, lobo medio basi profunde tricarinato, apice integro obtuso ; 

calcare tenui brevi ; columna brevi crassa. 

There were no leaves with the specimen. The raceme is laxly 8-flowered on the 

uj)per part of the leafless peduncle, which is 12-20 inches long. Sepals much smaller 

than the petals, the lateral 4 lines long by 2 J lines wide, the dorsal narrower (IJ line 

wide), the petals 7 lines long and broad. The lateral lobes of the labellum form the 

sides of a deep sac, of which the spur is a slender pointed continuation 2 lines long ; 

the median lobe is 4 J lines long, the lower part convex and deeply tricarinate on the 

crest, with raised lateral veins below ; the apical lobe is more flattened and spreading. 

The short thick column is 2 lines long. 

Rab. Milanji. 

There is in Herb. Kew. a specimen (Livingstone's Zambesi Expedition, Manganja 
Hills), with somewhat larger flowers, slightly narrower longer sepals, a blunt thicker 
spur, and apparently of a lighter colour, but the characters of the lip are the same, and 1 
doubt whether it can be separated from the Milanji plant. 

Brachycorythis pleistophylla, Eeichb. f. Otia Bot. Hamb. p. 104. 

Rab. Livingstone's S. Afr. Exped., Moramballa, Dec. 1838 (C. Meller ; Herb. Kew.); 
Nyasa-land, 1891 {Buchanan, no. 1345 ; Herb. Kew.) ; Milanji, Oct. 1891 {A. Whyte). 

The flowers of the Milanji specimen are larger than those of the type, and the bracts, 
differ in being prolonged into a subulate aristate apex half as long as the leaf. In 
Buchanan's specimen also the bracts have this character, while the flowers are inter- 
mediate in size. 

POLYSTACHYA, Sp. n. 

Rab. Milanji, Oct. 

Material insufficient for description. 

In Herb. Kew. there is a specimen with ripe fruit, evidently of the same species, 
collected by Sir J. Kirk at Zomba, Zambesia, (5000-7000 ft., in Sept. 1859 ; that is, in 
the same district and at about the same altitude as the Milanji plant. 



COLLECTED BY MR. ALEXANDER WJIYTE. 47 

The plants turn black on drying; the slightly ribbed leaf-sheaths decrease in length 
from 14 lines at the base of the stem to 10 lines just below the siini)lc spike [a^^U 
blade linear-oblong, with a rounded apex and fine prominent longitudinal veins 2| to 
3|in. long, 6-7 lines wide. The acuminate bracts are about 3 lines long, the sf)ike 
2 j in. The flowers, now dropped, have been few and distant. 

DisA zoMBAENSis, Rendlc, sp. n. (Plate VII. A, figs. 1-3.) Insignis caule robusto, 
rufescente, infra foliato, foliis quinque spiraliter ordinatis, rufescentibus, late 
lanceolatis, subacuminatis, sessilibus, supra in spicam densam evoluto; bracteis 
lanceolatis, foliis consimilibus, sed minoribus, infra flores sessiles excedentibus, supra 
iis subsequalibus aut apice brevioribus ; sepalis lateralibus ovali-obh)ngis, apice 
oblique apiculatis, venis tribus conspicuis converge ntibus, scpalo dorsali majore 
galeato in calcar longum, tenue, subrecurvatum, seraipollicare producto ; petalis 
ligulatis, galea inclusis, semipollicaribus ; labello tenui, filif(tnni; columiia crassa, 
brevi, truncata; ovario hirto, subcylindrico, utrinque attcnuato. 
The stem is 2 ft. 4 in. long ; the lower 7 in. are leafy, with broadly lanceolate leaves, 
decreasing in size upwards from 3 J to 3 in. long by 1-f in. broad ; the remainder forms 
a dense spike of sessile flowers, subtended by lanceolate bracts smaller than, but otlKT- 
wise similar to, the leaves, and also decreasing upwards, the lower ones exceeding 
the flowers, the uppermost falling short of them. The oval-oblong sepals are 6-7 lines 
long by 3-3 J wide, with three fine but well-marked darker longitudinal veins converging 
at the apex, and several fainter ones between these and the margin; the large hooded 
posterior sepal is f in. high by J in. broad, and produced into a narrow, slightly recurved 
dorsal spur, J in. long. The ligulate petals are 6 lines long by 1 broad, 5-nerved below, 
3-nerved above. The filiform lip is 5 lines long by J line wide. The ovary is j in. long 
and 2 J lines thick, subcylindrical, narrowing at either end. 
ffab. Zomba, Dec. 1891. No. 3. 
Resembles Disa Trailer i, but the flowers are twice the size. 

DiSA (§ Herschelia) hamatopetala, Rendle, sp. n. (Plate VII. B, figs. 4-6.) Erecta, 
glabra, foliis radicalibus anguste gramineis ; scapo flexuoso, laxe 2-7-floro, bracteis 
brunneis, membranaceis, ovatis acuminatis, sterilibus, margine sub medio connatis ; 
floribus breviter pedicellatis ; sepalis lateralibus ovali-oblongis, obtusis, dorsali 
galeato, calcare brevi recto-cylindrico; petalis basi columnse adnatis, hamatis, apice 
late emarginatis ; labello elongate, triangulari, piano, supra basin profunde fimbriato. 
The flexuose stem is densely surrounded at the base by the remains of the withered 
grass-like leaves, and varies in length from 6 or 7 inches to more than afoot, the shorter 
forms bearing only one or two flowers, which are also smaller than those on the larger, 
more robust stems. The bracts on the lower part of the flower-stem have their edges con- 
nate for half their length or more ; in the region of the flowers they are free. The larger 
flowers measure 1^ in. from the top of the si^ur to the end of the lateral sepals ; some of 



48 THE PLAINTS OE MlLANJl, NTASA-LAND, 

the smaller ones are less than 1 in. (11 lines). In the larger the lateral sepals are | in. 
long by J in. broad, the hooded dorsal one J in. high by nearly J in. from front to back ; 
the spur is 2 lines long by IJ in diameter. The hooked petals are nearly J in. long, 
laterally attached by the broader base to the side of the column, widening again near the 
broadly emarginate apex. Lip f in. long, tapering upwards, the sides, except near the 
base, deeply fimbriate as well as the apex. 

Rab. Milanji, 6000 ft., Oct. 1891 {A. Whyte, nos. 36, 136, & 199); Nyasa-land, 
1891 {Buchanan, no. 1016) ; Higher Plateau, North of L. Nyasa, Oct. 1880 (Thomson) ; 
Zomba, Zambesia, 6000-7000 ft., Sept. 1859 (Kirk) ; between L. Tanganyika and 
Nyasa, April 1890 {M. H. Johnston). All in Herb. Kew. 

Belongs to the section Hei^schelia, and is near Bisa lacera, Sw., var. ^nultijida, N, E. 
Brown, figured in the Bot. Mag. 7066 , bat is distinguished by the blunt sepals, the 
characteristic shape of the petals, and the blunter spur in a straight line with the base 
of the hooded sepal. The lip, moreover, is much more deeply fimbriated. 

Aristea Johnstoniana, Ilendle, sp. n. Herba foliis radicalibus subpedalibus; rigidis, 

linearibus, superne attenuatis, acutis ; pedunculis pedalibus, complanatis, aphyllis, 

vix alatis ; braetea vaginante acuminata, nunc inflorescentiam paucifloram superante, 

nunc breviore ; spathge valva exteriore triangulari acuta, interiore brunnea, membra- 

nacea, baud lacerata ; perianthii segmentis obovatis, patentibus. 

The leaves, which are condensed into a radical rosette, are nearly a foot long and 

1-2 lines broad. The very narrowly winged leafless peduncle ends in a small, usually 

3-flowered stalked head above a green sheathing acuminate bract, which sometimes 

overtops the head, sometimes falls short of its stalk ; there is usually a smaller lateral 

cluster sessile in the axil of the bract and included by it. The outer spathe- valve is 

green, with a broad, brown, membranous margin, and 5-6 lines long; the inner brown, 

membranous. The perianth-segments of the outer whorl are 4 lines long by 1^ broad, 

with a dark-coloured midrib projecting as an apiculate apex ; those of the inner whorl 

are broader (2 lines), with a rounded apex. Capsule not seen. 

Hal). Milanji, 6000 ft., Oct. Nos. 14 & 81. 

Near A. alata. Baker, from the mountains of the Masai country, but easily distin- 
guished by its very narrowly winged leafless peduncle ; it is also allied to A. abyssinica. 
Pax, but the latter has the broadly winged peduncle of A. alata and less rigid leaves, 
while the inflorescence is either a single head, or, if a second small axillary cluster is 
present, the whole is overtopped by a large bract *. 

* Ahihtea Tatxoeiana, Rendle, sp. n, Herba foliis 3-6-pollicaribus, radicalibus, rigidis, linearibus, superne attenu- 
atis, acutis vel subpungentibus ; pedunculis 4-8-pollicaribus, complauatis, aphyllis, angusto alatis; inflorescentia 
A. Johnstoniance, sed braetea semper breviore ; perianthii segmentis suberectis, fere spathulatis ; capsula obloiiga 
rotuudata, breviter pedicellata. 
The leaves closely resemble those of A. Johnstoniana, but are shorter, being 3-0 in. long and I3 to nearly '2 lines 
broad : they are generally shorter than the leafless peduncle, which is 4-8 iu. long, about 1 line broad, and narrowly 
winged. The inflorescence resembles that of A. Johnstoniana, but the bract always falls short of it ; the triangular- 
acute outer spathes are 5 lines long. The suberect perianth-segments are a vivid blue in colour, almost spathulate 



COLLECTED BT MR. ALEXANDER WIITTE. 



W) 



DiERAMA PENDULA, Baker, in Journ. Linn. Soc. xvi. p. 99. 

Sab. Milanji, Oct.; South-eastern provinces of Cape Colony, and northward to Natal 
and Mt. Kilimanjaro. 

Gladiolus (§ Eugladiolus) Melleri, Baker, in Journ. Bot. 187(), p. tiU. 

Apparently a fine form ; the flowers and capsules are considerably larger than tlie 
type, but I do not think it can be separated as a distinct species. 

Sab. Milanji, 6000 ft., Oct. 1891 {A. WJiyte, nos. 171, 195); Zambesi Highlands 
{Meller & Kirk; Herb. Kew.); Namuli, Makua Country {J. T. Last; Herb! Kew; 
Nyasa-land, 1891 [Buchanan, no. 1333 ; Herb. Kew.). 

Gladiolus (§ Hebea) Oatesii, Rolfe, in Gates, ' Matabele-land,' edit. 2. 

Sab. Milanji, Oct. 1891 (A. Whyte); Matabele-land (Oaies) ; Zambesi Highlands 
(Buchanan, no. 464) ; Transvaal [Holiib). 

Cyrtanthus Welwitschii, Hiern, ex Baker, in Journ. Bot. 1878, p. 197. 

A smaller plant than the type, but evidently the same species. A very Miiiil:ii- 
specimen was recently collected by Buchanan. 

Sab. Milanji, 6000ft., Oct. 1891 {A. Whyte, nos. 130 8l 149); Huilla {Wclwilsch, 
no. 4028) ; Nyasa-land (Buchanan, no. 993). 

Hypoxis angustifolia, Lam. Encyc. iii. p. 182. 

Sab. Milanji, 6000 ft., Oct., no. 119); KafFraria; Orange Free State; Tropical Afriwi 
(Angola, Moramballa, Zanzibar) ; Madagascar, Mauritius, and Bourbon. 

Vellozia (§ Xerophyta) sflendens, Eendle, sp. n. (Plate VIII.) Caulescens, sesqui- 

pedalis, caudicibus crassis, valide tunicatis, simplicibus vel furcatis ; foliis rosulatis, 

rigidis, vagina brunnea 2-3-pollicari, lamina lineari-subulata, complanata, glabra, 

margine scabrida; floribus solitariis in axillis bractearum ensiformium, quani 

pedicelli nigricantes scabridi vix breviorum ; perianthii lobis ovati-lanceolatis, apice 

acuminatis, falcatis ; antheris sessilibus ligulatis ; ovario turbinato, velut peduuculi 

scabrido. 

Caulescent, about 1 J ft. high. Caudices f -1^ in. thick, with a woody cylinder 4-6 lines 

thick, the rest composed of a mass of imbricating brown sheaths with strong longitudinal 

ribs ; simple or forked. Leaves about six in number, rosulate on the top of the caudex 



in shape, 5 lines long by 1| broad above the middle ; those of the outer whorl are slightly narrower, with a thick, 
dark-coloured midrib, forming a slightly apiculate apex. The oblong capsule is ^ in. long, on a shorter pedicel. 

Hab. Mt. Kilimanjaro, 7000 ft., 1884 (If. H. Johnston, no. 14) ; higher slopes up to 10,000 ft. above Morang 
{Rev. W. E. Taylor, 1888 ; Herb. Mus. Erit.). 

Johnston's specimens are referred in Baker's * Monograph of Iridese," p. 141, to Aristea alata, as a dwarfed form 
of the species, but are distinguished by the leafless and narrowly winged peduncle. A. Tayloriami is a smaller 
plant than A. Johnstoniana, from which it also differs in the shape of its perianth-lobes, which are longer, 
narrower, and more spathulate. 

SECOND series. — BOTANY, VOL. IV. H 



Mo. «-t. G. 
1393. 



bO THE PLAJVTS OF MILANJI, NTASA-LAND, 

or its branches, rigid ; sheath brown, 2-3 inches long ; blade of the outer leaves 1-1 J ft. 
long, 5-7 lines wide near the base, gradually tapering to a filiform apex ; the inner are 
proportionately smaller, flattened, glabrous, with a scabrous margin, toothed in the young 
leaf, and a midrib prominent on both surfaces. Peduncle single-flowered in the axil 
of ensiform, slightly shorter bracts ; 6-9 in. long, |^-f line thick, thickly covered with short, 
dark emergences. Perianth-cup very short, fleshy, and covered with protuberances like 
the ovary. Perianth-limbs ovate-lanceolate, 2J-2J in. long by 6-9 lines wide ; there is 
a small scale at the base on either side the insertion of the anther ; apex acuminate and 
falcate ; the base of those of the outer whorl covered with black emergences like those on 
the short perianth-tube, which are also continued up the central veins nearly to the apex, 
diminishing upwards in size and number. Anthers sessile, inserted on the perianth-limb 
about 1 line above their base, ligulate, 9 lines long ; connective slightly prolonged (^ line) 
beyond the anther. Ovary turbinate, J in. long, clothed with emergences like the 
peduncle. Style 1| in. long. 

Mab. Milanji. 

The short caudices bear magnificent clusters of flowers, which are larger than in any 
species hitherto described. 

Smilax Kraussiana, Meisn. in Flora, 1845, p. 312. 

Sab. Milanji, 6000 ft., Oct., nos. 119 & 152 ; Natal ; East Tropical Africa ; Sierra 
Leone ; Cape Verd. 

Asparagus virgattjs. Baker, in Saund. Refug. Bot. t. 214. 

Hab. Milanji, Oct. ; Cape of Good Hope ; Natal ; East Tropical Africa {Bev. W. 
E. Taylor, Herb. Mus. Brit.), Morang and between Zanzibar and Uyui; Maungu 
[S. jff. Johnston ; Kilimanjaro Exped.). 

Asparagus plumosus, Baker, in Journ. Linn. Soc. xiv. p. 613. 

Hab. Milanji, Oct. ; Cape of Good Hope ; Kaffraria; Natal; Nyasa-land {Buchanan, 
nos. 750 & 1470). 

Htlonome reticulata, Webb, Phyt. Can. iii. p. 320. 

Sab. Milanji Plateau, Oct., no. 60 ; Cape of Good Hope ; Natal. 

Drimia robusta, Baker, Saund. Refug. Bot. t. 190. 

Sab. Milanji, Oct., no. 6; Cape of Good Hope; East Griqualand {W. Tyson, 
no. 1108). 

TJrginea MARiTiMA, Baker, Journ. Linn. Soc. xiii. p. 221. 

Sab. Milanji, Oct., no. 192 ; Canary Islands ; Mediterranean ; S. Africa. 

TJrginea nyas^, Bendle, sp. n. Herba bulbo parvo ; foliis hysteranthis ignotis ; scapo 
glabro, subcompresso, ssepe flexuoso, inf erne purpureo, 7-8-pollicari ; racemo 



COLLECTED BY MK. ALEXANDEE WHYTE. 51 

angusto, multifloro (25-50), 2-6-pollicari ; bracteis niinutis orbiculari-deltoideis, 
superioribus breviter, inferioribus longe calcaratis ; pedicellis asccndontibiis ; 
perianthii albidi segmentis lanceolatis, uninerviis ; stamininn filaraentis linoari- 
subulatis, merabranaeeis ; stylo filiformi; capsula ovoidea, seniiiiibiis in loculo 5-10. 
complanatis, testa atra. 
Only part of the evidently small bulb present in the specimen. The larger scapes are 
1 line in width ; the narrow racemes only attain a width of | in. when in fruit ; the 
ascending pedicels are 2-3 lines long. The short membranous bracts have in the upper 
part of the raceme short spurs in the middle of the back ; in the lower part these spurs 
become much longer, as much as 2 lines. The perianth, which is 2 J lines long, l)ccomes 
detached at the base, and is thrown off like a cap from the half-ripe fruit. The filameuts 
are thin, membranous, and 1 line long, the anthers J line. The filiform style is 1 line 
long. The green ovoid capsule is 2 J lines long by 1 J broad. The seeds are black and 
flattened. 

Eab. Milanji, 6000 ft., Oct. 1891 {A. Whyte) ; Nyasa-land, 1891 (Buchanan). 

Ornithogalum Eckloni, Schlecht. Linnaia, xxv. p. 177. 

Hab. Milanji, Oct. ; Cape of Good Hope ; Natal ; Tropical East Africa. 

Anthericum, sp. 

Plower-buds unopened. 
Hab. Milanji, Oct. No. 175. 

Anthericum (§Phalangium) milanjiantjm, Eendle, sp. n. Herba foliis plicatis, late 
linearibus, utrinque attenuatis, subacutis, s£epius falcatis, basi vaginantibus, subrigidis, 
papyraceis, multinerviis, margine scabridulis ; caule glabro, conipresso, subpedali ; 
bractea ima subulata ; panicula angusta, 3-4-pollicari, nondum evoluta ; bracteis 
membranaceis, subulatis ; pedicellis medio articulatis; perianthii segmentis lanceolatis, 
albis, 7-nerviis ; filamentis filiformibus, glabris ; stylo filiformi; capsulo elliptico, 
seminibus in loculo pluribus superpositis turgidis. 
The broadly linear leaves are folded at the midrib, which is thickened like the 
margins, 8-18 in. long, 7-12 lines broad, wdth from 36 to 52 longitudinal veins. The 
flowerino" stem is sheathed at the base by the leaves, and, including the inflorescence, 
is 12-15 in. long by 2-2f lines broad, with one subulate bract about halfway up. The 
inflorescence is a narrow panicle 3-4 in. long, the lower subulate bracts j in. long ; the 
lateral branches are short and few-flow^ered. The flowers were unopened, except one or 
two on a smaller inflorescence, where there was also one fruit. The perianth-segments 
are 4^ lines long, white, with 7 delicate green nerves ; the filaments are 1| lines long, the 
laro-e yellow anthers 2J lines. The filiform style is 4 lines long. The single elliptical 
capsule is 3 lines long, and contains several black turgid seeds. 
Sab. Milanji, Oct. No. 123. 

H 2 



-52 THE PLANTS OF MILANJI, NYASA-LAND, 

Anthericum (§Phalangium) nyas^, Rendle, sp. n. Foliis radicalibus plicatis, linearibus, 
acuminatis, apice pungentibus, rigidis, glabris, margine ciliolatis, lO-nerviis, quaiu 
scapus multo brevioribus ; scapo glabro, firmo, subterete, infra pauiculam semipedali 
(vel minus), bibracteato ; bracteis longe acuminatis; panicula laxa, ramis paucis 
ascendentibus ; bracteis subulatis, membranaceis ; pedieellis ascendentibus, medio 
articulatis, inferioribus ternis ; perianthii segmentis lanceolatis, albidis, semi- 
pollicaribus, 3-nerviis ; filamentis quam segmenta plus duplo minoribus, filiformibus, 
glabris ; antheris viridibus ; stylo filiformi ; capsulis globosis, speciminibus nostris 
immaturis, seminibus in loculo pluribus. 
The stiff linear leaves vary in size up to 9 in. in length and IJ lines in breadth ; the 
firm glabrous scape is 6 in. long or less ; the panicle is generally 8 in. long, and has 2-4 
distant, ascending, laxly -flowered branches, 2-4 in. long; the pedicels are in groups 
of 3 in the lower part and 3-4 lines long; the white perianth- segments are 5-6 lines 
long and IJ lines broad, and have 3 well-marked green dorsal veins ; the stamens are 
3 lines long, the anthers 1 line, the filiform style 3^-4 lines ; the few immature globose 
capsules 2-2J lines long. 
Hah. Milanji, 6000 ft., Oct. 

Chlorophttum blepharophylltjm, Schweinf. ex Baker, Journ. linn. Soc. xv. p. 327. 
Hah. Zomba, Sept. 1891 {A. Whyte) ; Gallabat (Schweinfurth) ; Transvaal (Barnes), 

CoMMELiNA AFRiCANA, Linn. Sp. PL p. 41. 

Hab. Zomba, Sept. 1891 (A. Whyte) ; S. Africa, Angola, Ins. Bourbon 

€oMMELiNA Bainesii, C. B. Clarke, Monogr. Phaner. iii. p. 184 ; var. glabrata, Eendle, 
var. nov. 

Differs from the type in the almost glabrous character of the leaves and the slight 
pubescence of the inter nodes ; the whole plant of C. Bainesii is described as " villoso- 
pubescens." 

Sab. Zomba, Sept. 1891. 

C. Bainesii occurs in S. Africa, Transvaal, Bechuanaland, and Angola. 

CoMMELiNA ALBESCENS, Hassk. ; Schwcinf . jEthiop. p. 210. 

Sab. Milanji, 6000 ft., Oct., No. 112 ; Abyssinia, Arabia, Beloochistan, Scinde. 

Anellema sinicum, Lindl. Bot. Reg. t. 659. 

Hah. Milanji Station, Oct., No. 39 ; S. Africa, Angola, Madagascar, E. India, Java, 
China. 

Aneilema ^quinoctiale, Kunth, Enum. iv. p. 72. 

Hah. Milanji, Oct., No. 131; Zomba, Sept.; Abyssinia, Congo, Loango, Angola; 
Arabia. 



COLLECTED BY MR. ALEXANDER WHYTE. 53 

J'loscopa glomerata, Ilassk. Commel. Iiid. p. 166. 

Hab. Zomba, Sept. ; S. Africa, from Cape Colony to tlie Zambesi ; Angola, Madagascar. 

Eriocaulon Sonderianum, Korn. in Linngea, xxvii. p. 669. 
Hab. Milanji, 6000 ft., Oct. (no. 115) ; Cape of Good Hope. 

Pycreus spissiplorus, C. B. Clarke, sp. n. Mediocris, umbella simplici aut in caput 
unicnm contracta; spiculis spicatis, lanceolatis, compressis, multifloris, fusee luteo- 
rubescentibus ; glumis spisse sitis, ovatis obtusis ; stylo 2-fido. 

Culms about 12 in. long; rhizome woody, lower sheatbs blackened and torn. Leaves 
f the length of the culm, If line broad. Radii of the umbel sometimes reaching 1| in. 
in length, more often short or almost absent ; bracts 3-5, reaching a length of 2J-3 in., 
spreading like the leaves. Spikelets resembling those of P. lanceiis, C. B. Clarke, in 
shape and size, but with as many as 16-20 flowers. Nut immature. 

Near JP. lanceus, C. B. Clarke, from which it differs in having the glumes much more 
crowded and a darker colour. The immature spikelets reciill those of P. capillaria^ Nees, 
but the inflorescence is different, resembling that of P. lanceus. 

Hal). Milanji, Sept. 

Pycreus nigricans, C. B. Clarke, ined. ; Cyperus nigricans, Steud. in Flora, 1812, p. 584. 
Hab. Milanji, 6000 ft., Oct. ; Abyssinia, Intschatkak {Schimper, ii. no. 1378) ; Gafat, 
8700 ft. {Schimper, no. 1286). 

Cyperus ochrocephalus, C. B. Clarke, ined. ; Rhynchospora ochrocephala, Boeck. in 
Flora, 1879, p. 568. 
Hab. Milanji, Sept.; Tropical Africa, Madagascar, and S. Africa (C sphcerocephalits, 
Vahl). 

Cyperus zambesiensis, C. B. Clarke, ined., in Herb, Kew. 
Hab. Milanji, Oct.; Buchanan, Nyasa-land (no. 647). 

Mariscus cf. DACTYLiEORMis, C. B. Clarke, ined.; Cyperus dacfyliformis, Boeck. Cyperac. 
p. 275. 
Hab. Milanji, Oct. ; Cape of Good Hope. 
The specimens are too young for more definite naming. 

Mariscus Sieberianus, Nees, var. e. polyphylla, C. B. Clarke, ined., in Herb. Kew. 
{M. polyphyllus, Steud.) 

Hab. Milanji, Oct. ; Abyssinia {Schimper, no. 1124); E. Africa, between Zanzibar and 
Uyui (Bev. W. E. Taylor, Herb. Mus. Brit.). 

This species is widely distributed through India, Malaya to Hong-kong, Formosa, and 
the Philippine Is., and occurs also in the Mascarene Is. and at the Cape of Good Hope. 

Varieties are found in St. Helena and Samoa. 



,54 THE PLANTS OF MILAN JI, NYASA-LAND, 

BuiiBOSTYLis SCHCENOIDES, C. B. Clarke, ined. Isolepis schoenoides^ Kunth. Enum. ii. p. 208. 
Sah. Milaiiji, 6000 ft., Oct., ^""o. 75; South Africa ; Tropical Africa, Angola ( Welwitsch), 
and Camaroons, 8000-10,000 ft. (S. II. Johnston). 

BuLBOSTTLis ORITUEPHES, C, B. Clarke, ined. Fimhristylis oritre flies, Bidley, Trans. Linn.. 
Soc. ser. 2 (Botany), ii. p. 155. 
Kah. Milanji, Oct. ; Angola ( Welwitsch). 

SciRPUs cosTATUs, Bocck. Cyperac. p. 457. 
Hah. Milanji, Sept. ; Abyssinia {Schimper), 

FuiRENA Welwitschii, Bidl. I. c. p. 161. 
Sab. Milanji, Oct. ; Angola ( JVelwitsch). 

AscoLEPis CAPENSis, Eidlcy, /. c. p. 164 ; Flatylepis capensis, Kth. Enum. ii. p. 269. 
Hab. Milanji, Oct. ; South Africa ; Tropical Africa { Welwitsch). 

Eriospora villosijla, C. B. Clarke, sp. n. Herba culmis longis, gracilibus, tenuiter 
pilosis ; foliis lineari-setaceis, vaginis fissis ; panicula laxa, ssepe composita ; bractea 
ima niagis pilosa, immo villosa ; spicis omnibus bisexualibus ; stylo 3-fido ; setis 
hypogynis, rigidiusculis, setuloso-scabris. 
The culms are a foot or more long, ^ line in diameter, and subterete. The leaves 
about 9 in. long and 1^ lines wide, with the margins involute when dry. The panicle 
is 5-7 in. long, with 15-25 spikes ; the fasciculed peduncles are as much as l-|-3 in. 
long below ; the pilose bract almost equals the panicle in length, and is sometimes 
almost hirsute. The spikelets are those of the genus. 
Mab. Milanji, 6000 ft., Oct. 

Near ^. Mekmanniana, C. B. Clarke, ined. in Herb. Kew., from the Transvaal, but easily 
distinguished, as the latter is glabrous and has not the lax panicle of the Milanji plant. 

Carex Wahlenbergiana, Boott, ? var. Schimperi, Boott. 

The species occurs in Bourbon, Mauritius, and Madagascar ; the variety, according to 
Engler's ' Hochgebirgsflora trop. Afrika,' in Abyssinia " bei Gerra um 2500"" " (Schimper, 
1863, no. 1314), and Eernando Po " um 2500™" {G. Mami ; ~C. Wahlenbergiana, fide 
Hook. fil. in Journ. Linn. Soc. vii. p. 225). 

It is allied with the wide-spread tropical Asiatic C. benyalensis, Boxb. 

Hab. Milanji, Oct. 

Carex Boryana, Schkuhr, Biedgr. ii. p. 43. 
A large form of the species. 
Hab. Milanji, 6000 ft., Oct. No. 85. 
A native of Bourbon and Mauritius ; found also in Eernando To, fide Hook. fil. /. c. 

Panicum (§Eupanicijm) pectinatijm, Bendle, sp. n. (Plate X. figs. 1-6.) Csespitosa, 
culmis tenuibus, subteretibus, glabris, foliosis, pedalibus et supra, glandulis pedi- 



COLLECTED BY MR. ALEXANDER WHTTE. 55 

cellatis super nodis pilosis suffultis ; foliorum vaginis longitudinaliter sulcatis, infor- 
ioribus scabrido-puhescentibus, supcriore glanduliforo ; ligula pilosa ; lamina linoari, 
apice attenuata, glabra ; paniculae ramis ascendentibus vol patcntibus, velut ramiilis 
pedicellisque glanduliferis ; spiculis ovalibus, subcorapressis, gluma 1"" brovi, ovata, 
3-pcctinata, purpurea, sublinea longa, 2'^'' lata, truncata, 5-nervia, pcctinata, 5-dontata, 
S*"" ovali-oblonga apice irregulariter dentata, 7-nervia, margine purpurea ; florifera 
quam prsecedens subbreviore, ovali, apice subobtusa et dorso sparse ciliolata ; palca 
quam gluma includente breviore, sub apice sparse pilosa. 
The slender leafy culms are longitudinally striate, 12-21 in. long, and be<ir a few 
long-stalked capitate hairs, evidently glandular, below the pilose nodes. The leaf-sheaths 
are longitudinally ridged, the lower ones scabrido-pubcscent, while the uppermost bears 
the stalked glands in the depressions between the ridges ; the ligule is represented by a 
row of pilose hairs at the mouth of the sheath ; the linear blade has a tapering, some- 
what cuspidate apex, is longitudinally striate with inconspicuous venation, 3-4 in. long, 
IJ lines broad, in the radical leafy shoots often with completely involute margin. The 
inflorescence is a panicle, with branches ascending or spreading, and giving an ovate 
general outline 3-5 in. long by 2-3 in. broad ; the axis, branches, and pedicels bear the 
characteristic glands, which are easily rubbed off the dry plant. The oval subcompressed 
spikelets are If lines long. The short outer glume is f line long, consisting in the upper 
part of three teeth, of which the middle is the longest ; the second is purple like the first, 
but larger, 1 J lines long, and 5-nerved, or with a shorter 6th nerve not reaching the top, 
with 5 apical teeth ; the third is oval oblong, If lines long, 7-nerved, with purplish edges 
and an irregularly toothed apex ; the 4th or flowering glume is very slightly shorter than 
the third, oval, flattened dorsally, with involute edges firmly enclosing the pale and flower : 
it is coriaceous with a hyaline margin, the upper third purple, the rest greenish, sparsely 
ciliolate on the back beneath the subobtuse apex. The pale is slightly shorter than the 
flowering glume, 1| lines long, flattened dorsally, with involute margin, subcoriaceous 
with hyaline edges^ glabrous except for a few hairs near the acute apex. The stigmas 
are plumose. Eruit not seen. 

Sab. Milanji, 6000 ft., Oct. No. 10. 

This grass is of interest from a biological standpoint owing to the long-stalked glands, 
which must be of importance in the life of the plant, while from a systematic point of 
view its pectinate glumes separate it from all other species of the genus Panicum. The 
latter character I was inclined to consider of generic value, but Prof. E. Hackel, whom I 
consulted, thinks the plant " a true Panicum, but somewhat anomalous in the genus on 
account of its toothed glumes." He also reminds me that, while no Panicum is known 
as yet which offers this character, a similar case occurs in the genus Muehlenhergia, 
where M. clomena, M, gracilis, and others have the second empty glume three-toothed, 
a fact which led Beauvois and Nuttall to separate them as distinct genera ; modern 
authors, however, including Bentham, have reunited them with Muehlenhergia. 

P. pectinatum falls in the section Eupanicum, its otherwise normal spikelets, with 



56 THE PLANTS OF MILAJVJl, NTASA-LAND, 

their slender pedicels, being laxly arranged on the branches of the panicle. It will form, 
however, as Hackel suggests, a new series comparable with the other seven into which 
Bentham divides the section. This series I propose to call JPectinatce. 

Pakicum (§Digitaiiia) MILAN jianum, Rcndle, sp. n. ; rhizomate tenui, forsan repente; 
culmo folioso, subbipedali, glabro, striato ; internodiis tribus infimis complanatis, 
superiore terete ; vagiuis glabris, ore pilosis, striatis ; ligula brevi, truncata, brunnea, 
membranacea ; lamina late lineari, superne attenuata, nervo prominente, scabrido- 
pilosa; spicis in axe brevi fastigiatis ; spiculisin rliacheos triangularis flexuosse late- 
ribus inferioribus paribus alternantibus ordinatis, subcompressis, unifloris ; pedicellis 
parium inaequalibus ; gluma l"""" minima acuta, 2*^* profunde incisa segmentis pilosis, 
^'^ maxima a tergo visa lanceolata, cum duabus carinis, pilosis, ^^ florifera late ovata 
apiculata, dorso rotundata, trinervia, glabra, coriacea, palea consimili, sed obtusa, 
inclusa ; antheris brunneis, stigmatibus plumosis. 
The apparently creeping slender rhizome is \\ lines in thickness ; the culm is nearly 
2 ft. long; the three lower internodes are compressed, the upper one terete above the 
leaf-sheath ; the nodes are brown and glabrous. The leaves are in a radical tuft and 
cauline; the sheaths of the former are 2J-3 in. long, of the latter 3 J, 5, and 9 in. 
respectively from below upwards ; the ligule is less than a line long, the broadly linear 
blade tapering above is about 6 in. long by 4-5 lines wide in the middle, with a con- 
spicuous midrib raised above, and numerous less conspicuous lateral veins, scabrido- 
pilose on both surfaces, but especially on the upper. The inflorescence consists of 
15 spikes arranged irregularly on a short axis 2 in. long; the spikes are 6 in. long, 
the spikelets, IJ lines long, are arranged in alternating pairs on the two lower sides 
of the triangular flexuose rhachis ; one pedicel of each pair is longer, sharply triangular, 
and closely denticulate like the rhachis, the shorter being terete, with a few small scattered 
teeth. The small acute (not short and rounded as in Digitaria sangulnalis^ Scop.) lowest 
glume is ^ hne long ; the second, deeply cut into hairy divisions, is 1 line ; the third and 
largest, appearing lanceolate from behind, has a prominent, minutely denticulate midrib 
on the flattened back, with two pilose lateral keels, and is 1^ lines long ; the broadly ovate 
flowering glume, with rounded back and thin transparent margin, encloses the similar 
but blunter and more delicate pale, which is 1 line long. 
Hab. Milanji, Oct. 

Closely allied to Digitaria commutata, Nees, but at once distinguished by the narrow, 
acute, projecting lowest glume, which in J), commutata is equilaterahy triangular and less 
projecting. The leaves are also broader. 

Sacchae,um (§ Eeiochrysis) puhpurattim, Eendle, sp. n. ; culmo glabro, nodis 3 ; 
vaginis duabus superioribus pubescentibus, infima glabra, innovationuni glabris \ 
ligula breviter ciliata ; lamina flexuosa, convoluta, aciculari vel subulata, pube- 
scente ; panicula parva, ovali, densa, pauciramea ; spiculis lanceolatis, sessilibus, 
quam pili fulvi involucrantes longioribus, pedicellatis cos subsequantibus ; gluma 



COLLECTED BY MR. ALEXANDER WHYTE. 57 

1™^ membranacea, ovato-lanceolata, suhohtusa, bicarinata, inargine inflexa, venis I- 
convergeiitibus, purpurea, marginibus, apice, et carinis pilis fulvis obtectis ; 
2*^* hyalina, late lanceolata, acutiuscula, l)icarinata, venis 3 converiientibus. mar- 
ginibus inflexis et carinis supra basin pilis fulvis eiliatis; 3''" hyalina, lanceolata, 
subacuta, enervia, margine inflexa, supra medium ciliata ; l^ quam 3''" duplo 
breviore, hyalina, subtrapezoidea, apice incisa, enervia, ciliata. 
The culms are 14-17 inches long, with three nodes, the uppermost naked for 21 inches 
below the panicle ; the lowest sheath on the culm is glabrous, the upper two are pubes- 
cent, those of the innovations glabrous ; the blades convolute, appearing acicular or 
subulate, and covered with a silky pubescence ; the lowest on the culm is 3 in. long, 
the upper two about 1 in. and J in. respectively, on the innovations 1-7 in. long; they 
are J line in diameter when rolled. The panicle is small, oval, dense, and but little 
branched, 1^-2 in. long by 6-8 lines broad ; the colour is tawny, tinged with wine- 
red. The sessile spikelets are 3 lines long by 1 broad, with an involucre of stiff tawny 
hairs about half the length of the outermost glume ; the stalked spikelets are 2J lines 
long, with a denser, longer, and more spreading involucre, slightly shorter than, or nearly 
equal to, the spikelet. The outermost glume is 3 lines long by 1 broad, with two stronger 
lateral veins parallel with the keels and two central ones which anastomose, one or botli 
falling short of the apex ; the inflexed margins, the apex, and the upper portion of 
the two keels bear tawny hairs, the rest of the surface being glabrous ; the colour is 
wine-red. The second glume is nearly 3 lines long by § line wide, with a midrib and 
two lateral veins ; the upper f of the keels and the infolded margin almost from the base 
bear tawny hairs ; the third glume is nearly 2i lines long, veinless ; the upper f of the 
inflexed margin and the upper half of the surface is hairy, but the hairs are shorter 
and less coloured than in the last ; two rows of hairs in the lower portion recall the 
keels of the outer glumes ; the fourth is about half the length of the third, is ciliated, 
and has a cut subtruncate apex. 

S^ab. Milanji, Oct. (no. 8) ; Nyasa-land, 1891 (Buchanan, no. 997). 
Allied to the Cape and West Tropical- African species S. 31unroamfm, Hack., but easily 
distinguished by the smaller inflorescence, the reddish and membranous (not coriaceous) 
glumes of the sessile floret, and the shorter paler tufts of hair. 

Haepechloa altera, Rendle, sp. n. (Plate X. figs. 7-12.) Percnnis, ca3spitosa, 
rhizomate duro ; culmis tenuibus, foliosis, vaginis fere obtectis, qua nudi pubes- 
centibus, erectis, complanatis, flexuosis, quam folia brevioribus; vaginis longis, 
Ijevibus, plicatis, carinatis ; ligula brevi, pilosa ; lamina lineari, glabra, striata, 
flexuosa, plicata, valde carinata, margine involuta, apice pungente ; spica unilaterali 
vel disticha, falcata ; gluma 1°"* membranacea, complanata, scaplioidea, acutiuscula, 
uninervia ; 2^^* prsecedenti simili, sed breviore ; 3*** florifera, brevissime pedicellata, 
complanata, lanceolata, apice bifida, carina valida marginibusque dense eiliatis; 
palea inclusa, hyalina, profunde bicarinata ; glumis duabus superioribus minoribus, 
sterilibus, truncatis, apice emarginatis. 

SECOND SERIES. — BOTANY, VOL. IV. I 



58 THE PLANTS OF MILAN JI, NYASA-LAND, 

A perennial tufted grass with a hard rhizome ; the annual shoots surrounded at the 
hase with the dried persistent leaf -sheaths, of which the internal ones are covered more or 
less with a woolly puhescence. The slender leafy culm is almost completely covered by 
the leaf-sheaths, woolly pubescent where exposed, is flattened, erect, flexuose, overtopped 
by the radical leaves and those of the leafy shoots, 7-10 inches long. The leaf-sheaths 
are long and smooth ; the ligule is reduced to a ring of short hairs ; the linear flexuose 
blades have a strong keel, closely involute edges, and a pungent apex, those of the radical 
leaves are 8-9 inches long, that of the upper leaf of the culm 1-lJ inches. The inflores- 
cence is a falcate, unilateral, or sometimes distichous spike 1 J-2 inches long ; the spikelets 
are 2 J lines long, with one flower and two uj)per barren glumes. The outermost glume 
is membranous with hyaline edges, compressed and boat-shaped, when opened out narrow, 
triangular, with a blunt apex, light brown, with a dark keel and no lateral veins, 2 J lines 
long ; the second is somewhat similar, but slightly shorter and of a long oval shape ; the 
short rhachilla beneath the flowering glume bears short tufts of hairs. The folded flowering 
glume is lanceolate, when opened out ovate, with a bifid apex ; the sides are delicately 
membranous, with a purple tinge ; the strong keel and margins are densely ciliate, with 
straw-coloured hairs, 1^ lines long ; the included narrow hyaline pale is deeply bicarinate 
and slightly shorter. The two upper barren glumes are smaller, thin, hyaline, truncate, 
with an emarginate apex, about 1 line long, the lower oval-oblong, the upper narrower. 

Sah. Milanji, 6000 ft., Oct. 1891 {A. Whyte, no. 64); Nyasa-land, 1891 (Buchanan, 
no. 163) ; Transvaal, Aug. 1880 {Nelson, no. 14) ; Natal. 

Easily distinguished from jff. capensis, Kunth, the only species of the genus hitherto 
described, by its larger, narrower, and less rigid leaves exceeding the spike, which is also 
narrower and more delicate, the smaller spikelet, the near approach to equality in the 
two outermost glumes, the shape of the upper barren glumes, and the absence of a third 
staminate flower. 

Trichol^na ROSEA, Necs, El. Afr. Austr. i. p. 17. 
Sab. Milanji, Oct. ; S. Africa. 

Trichol^na aff. T. leucantha, Hochst. ex Steud. Gram. p. 92. 
No leaves ; specimen insufficient for determination. 
Sai. Milanji, Oct. 

Andropogon hirtus, Linn. Sp. PL p. 1046. 
Sab. Milanji, Oct. ; Europe, Asia, and Africa. 

KcELERiA CRisTATA, Pers. Syuops. i. p. 97. 

^a6. . Milanji, Oct.; Nyasa-land, 1891 {Buchanan, no. 991); Europe — Caucasus, 
Siberia. 

Eragrostis, cf. E. MINOR, Host, El. Austr. i. p. 135. 
Material insufficient for determination. 
Sah, Milanji, Oct. 



COLLECTED BT MR. ALEXANDER WHYTE. 59> 

Pestuca costata, Nees, Fl. Afr. Austr. i. p. 447. 
Eab. Milanji, Oct. ; S. Africa. 

Pestuca milanjiana, Rendle, sp. n. ; rhizomatc subvalido; culmo 1-2-peclali, siuiplici. 
erecto, compresso, laevi, folioso, vaginis radicalibus demum in fibras solutis ; ligula 
membranacca, alba, truncata, 1 lin. longa ; foliis linearibus utrinque attcnuatis, 
planis vel siccis ssepe convolutis, glabris, supra sulcatis, subtus sublawibus ; paniculye 
laxse substrictse ovalis semipedalis radiis geminis elongatis, ascendentibus, triquetris, 
Isevibus, prope medium divisis ; spiculis ovali-lanceolatis, valde comj)irssis, 4-floris ; 
flosculis dissitiusculis, baud coloratis; glurais lanceolatis, subscabiidis, sterilibus 
acutis, fertilibus 5-nerviis, apice bidente, carina in aristam intermediam pcrbrcvem 
producta. 
The leaves which overtop the panicle may be a foot or more in length (including the 
sheath, the length of the blade being from 6-9 inches) and from 1-1 J lines at the widest 
part ; the upper part is long and very attenuate. The panicle is from 5 to (> inches long ; 
the sterile glumes, which have two conspicuous nerves on each side (a third is rarely 
present), are 4 lines long and the lowest fertile one 5 lines. 
Sab. Milanji, Oct. 

Near the South- African species F. costata, Nees, and F. scahra^ Vahl ; distinguished 
from the former by the narrower panicle, more lanceolate spikclets, colourless glumes, 
and proportionately larger sterile glumes, these being nearly as long as the lower fertile 
ones ; from the latter by its lax panicle, more lanceolate spike, and less scabridity. The 
culm flattened throughout is also a distinguishing characteristic. 

Bromus MILAN jianus, B^endle, sp. n. ; culmo elongato, bipedal i, laevi, striato, compresso, 

folioso ; ligula membranacca, alba, truncata, demum fimbriata, 1 lin. longa; fohis 

elongatis ; late linearibus, utrinque attenuatis, planis, supra sparse villosis, subtus 

Isevibus, paniculae angustse gracilis semipedalis radiis multis verticillatis ternis, 

ascendentibus ; spiculis compressis, trifloris, lateralibus breviter, terminalibus longe 

pedicellatis ; glumis sterilibus lanceolatis acutis, apice breviter setigeris, inferiore 

uninervia, altera trinervia ; fertilibus glabris, late lanceolatis, inconspicue 7-nerv iis, 

apice profunde bifidis, carina scabridula in aristam flexuosam glumam parce 

excedentem producta. 

The acute tip of the uppermost leaf of the culm falls a little short of the top of the 

panicle. The leaf -blades may attain a length of 9 in., and are 2i lines wide at their 

broadest part. The graceful curviug panicle is 6-7 in., the spikelets are 5 lines long. 

The lowest sterile glume is 2^ lines long and very acute; the upper is broader, 3^ lines 

long and ^ line broad. The lowest fertile glume measures, without the awn, 5 by 

IJ lines. Rhachilla pilose beneath the florets. 

Hab. Milanji, Oct. No. 9. 

Very near the two closely related species from Central Madagascar, B. arrhenatheroides 
and B. avenmdes, described by Baker in Journ. Linn. Soc. xx. pp. 301, 302, being 

i2 



60 



THE PLANTS OP MILANJI, KTASA-LAND, 



remarkable for the deeply bifid flowering glumes and tbe avenoid habit. It is especially 
near the former, its spikelets being 3-flowered, but is distinguished by its smaller sterile 
glumes, slightly broader, very inconspicuously nerved fertile ones, and broader flattened 
leaf-blade. The awn, moreover, is not sharply geniculate, but flexuose. 

GYMNOSPERMJE. (By A. B. Eendle, M.A., P.L.S.) 

WiDDRiNGTONiA Whytei, Kendlc, sp. n. (Plate IX. figs. 6-10.) Arbor monoica; ramis 
et ramulis fastigiatis, illis teretibus, foliorum vestigiis ligneis notatis, his subteretibus 
cum squamis arete spiraliter imbricatis, infra adnatis, supra patentibus subpungenti- 




WlDDBINGTONIA "WhYTEI. 

From a photograph taken at ililanji, alt. 7500 ft., 12th Dec, 1893. 

bus ; squamis in ramulis ultimis decussatis, subquadratis, apice crassis ; strobilis 
masculis solitariis, terminalibus, ovalibus ; antheris 8, decussatis, orbiculari-ovalibus, 
bracteis geminis, subtendentibus ; foemineis 4-6 in ramulo brevi aggregatis, sub- 



COLLECTED BY MR. ALEXANDER WHYTE. CI 

globosis, squamis 4 basi connatis suba^qualibus, duabus paulo majoribiis, oblongis 
obtusis, dorso rugosis vel subtuberculatis, infra apiceiu umbone conico armatis. 

A magnificent tree, reaching a height of 140 feet, sometimes with a clear straiglit stem 
for 90 feet, and a diameter of 5J feet at G feet from the base, sometimes giving off 
long straggling branches nearer the base. The branclilets bearing the leafy shoots 
are rough with the woody remains of the squamiform leaves ; those of the next higher 
order have spirally-arranged imbricating scales, with a suboval lower part adnate to 
the stem and an upper free triangular part with a subpungent apex ; the ultimate 
branchlets have close-fitting decussate quadrate scales with a thickened apex ; a transverse 
section of these shows three longitudinal resin-canals traversing the interior of tlie tissue, 
a larger median and two smaller lateral ones. The male cones arc solitary, terminal, 
oval, and consist of eight orbicular-oval, subapiculate anthers, subtended by two spreading 
apiculate bracts. The female cones are crowded 4-6 together on a sliort lateral shoot, 
are subglobose, about | in. long and f-1 in. wide w hen open. Of the four scales two are 
slightly larger ; they are connate at the base, oblong-obtuse, more or less tuberculate on 
the back, with a conical spur below the tip ; the two internal faces are flat, subconcave 
below, and smooth or longitudinally striated. Numerous hilar scars round the base of 
the scales indicate the position of the winged seeds. 

The timber is of a pale reddish colour, of excellent quality, and easily worked. Tlio 
bark on old trees is of great thickness, consisting of layers annually shed and renewed. 

According to Mr. Whyte these fine trees are rapidly disappearing before the forest fires, 
the few^ left being confined to the upper ravines and valleys, the largest forest finding a 
comparatively secure habitat in the damp gorges of the Lutshenya valley. 

It is of interest to note that the most nearly allied species, W. juniperoides, Endl., seems 
also to be dying out in the Cederberg Mts., for Parlatore says : — " Olim vastas sylvas 
efformabat, nunc minus obvia." 

Sab. Milanji, 6000-8000 ft. 

Mr. Whyte's discovery extends the range of the genus into Tropical Africa. 

Closely allied to W. juniperoides of the Cederberg Mts., Cape Colony, but distinguished 
by its smaller fruits, with much less tuberculated scales, and also by tlie shape and internal 
structure of the leaf-scales on the ultimate branchlets. In TF. Whijtei these are quadrate, 
with a slightly larger median and two lateral longitudinal resin-canals ; while in the 
Cederberg plant they are more rounded and have one very large canal traversing the 

centre. 

Bentham and Hooker regard Widdringtonia as a section of Callitris, a genus made up 
of the North-African C. qiiadrivalvis, Vent., the Australian genus Frenela, and the 
S.- African and Mascarene Widdringtonia; Dr. Masters, however, in a recent paper 
(Journ. Linn. Soc. xxx. p. 16) restores Widdinngtonia^ and I am glad to have the oppor- 
tunity of keeping it up. 

PoDOCAEPUS MILANJIANA, Rcndlc, sp. H. ? Arbuscula, foliis confertis lineari-lanceolatis, 
acutis, submucronatis, basi in petiolum perbrevem angustatis et saepissime tortis, 



62 THE PLANTS OF MILANJI, NYASA-LAND, 

coriaceis, rectis vel subfalcatis, nervo longitudinal! prominente notatis, margine in 

sicco vix revolutis ; amentis masculis 2-3-aggregatis, axillaribus, sessilibus, oblongo- 

cylindraceis, obtusis, crassiusculis ; squamis antheriferis late ovatis, apice rotunde 

dilatatis ; pedicellis fructiferis solitariis, axillaribus, 1-2-floris. 

The somewhat crowded leaves are almost sessile, with stomata on the lower surface only, 

2-3 inches long, 3-5 lines broad ; those of the male specimens are slightly broader than 

those of the female. The male spikes are 1 inch long by 2 lines in diameter, the fruiting 

pedicels 4-5 lines long. Only very young fruits were present. 

Near Fodocarpus Thunhergii, Hook, fil., from the Cape, but distinguished by its 
narrower, longer leaves, and much larger, not solitary spikes. It differs from the alhed 
P. elongata^ L'Herit., in its much larger leaves and the form of the male spikes, which 
are twice as thick, but shorter. 

Sah. Milanji, 6000 ft., Oct. Nos. 34 & 39. 



MOSSES. (By A. Gepp, M.A., P.L.S.) 

fragments of Dicranum, Letjcoloma, Bryijm, and Thuidium. 

PoLYTRicHUM COMMUNE, Linn. Sp. PI. p. 1109. 
Widely distributed. 

Aerobryum (§ Eriocladitjm) capbnse, C. Muell.in Linnsea, xl. p. 262. 
Distrib. S. Africa. 

PiLOTRiCHELLA (§ Orthostichella) imbricata, Jacg. Genera et Species Muscorum,- 
ii. p. 159. 
JDistrih. E. and W. Tropical Africa. 

PlIiOTRICHELLA (§ OrTHOSTICHELLA) PSErDO-IMBRICATA, Jaeg. loc. cU. 

The Milanji specimens agree closely with a form in Herb. Hampe collected by Hilde- 
brandt in the Comoro Islands, and bearing the MS. name Neckera {Orthostichella) 
decolo7^ans, Hampe. 

PoROTRiCHUM DENTATUM, Gepp, sp. n. Dioicum (?). ithizoma repens. Caules ascen- 
dentes, inferne simplices, foliis sparse vestiti, pollicares, superne flabellati. Plabellum 
lutescens, complanatum, laxe dendroideum, 1-2-pollicare. E^ami pollicares, subsim- 
plices vel irregulariter pinnatim ramosi. Polia caulina e basi lata (1 mm.), amplexi- 
cauli, erecto -patent ia, deltoideo-ovata, breviter acuminata, baud plicata ; ramea e basi 
angustiore ad alas impressa, patula ovata vel ovato-lanceolata, caviuscula, plerumque 
plica medio exarata, apice obtuso apiculato, margine late recurvo, inferne serrulato, 
superne serrate, versus apice m dentato, nervo valido ad | evanido, cellulis inferioribus 
linearibus ad alas impressas brevibus incrassatis, superioribus ovali-rotundatis. 



COLLECTED BY MR. ALEXANDER WHTTE. 68 

Polia perichsetialia pallida, inferne lutea, ovata, longe acuminata, ecostata, margine 
serrulate. Archegonia circa 25, cum pallidis paraphysibus dense conferta. 
This plant somewhat resembles Forotrichum pentKeforme, C. Muell.,in size and general 
appearance, but differs in the less crowded, more ascending branches, and the areolation 
of the leaves. 

Male flowers not observed. Rhizome 0*25 mm. thick, angulate. Flabellate stems 
5-7 cm. long. Stem-leaves 1"5 mm. long; areolation at base lax, irregularly oval, linear 
above, and towards apex oval, with thick cell-walls sparsely pitted. Branch with its 
leaves about 1 mm. thick. Branch-leaves 1'25-1'5 mm. long, not complanate normally ; 
nerve often concealed in a deep fold. Perichaetial leaves with cells crowded, hexagonal, 
thin-walled at base ; marginal row quadrate, very delicate, linear above, with walls thick, 
gelatinoid, pitted. 
Hah. Milanji. 

HEPATICS. (By A. Gepp, M.A., P.L.S.) 

Plagiochila Rtttenbergii, Gottsche, in Abhandl. naturwiss. Vereines Bremen, vii. 
p. 338. 

The Milanji specimens agree in most respects with Gottsche's description and figures. 
They differ in having the ala of the perianth in some cases ciliato-dentate (4 cilia) towards 
the apex. 

Distrib. Madagascar. 

LOPHOCOLEA, sp. 

Eadtjla, sp. Near R. Boryana, Nees. 

Phragmicoma Pappeana, Nees, Synops. Hepat. p. 296. 
Distrib. S. and E. Africa. 

XeJEUNEA (§ LOPHOLEJErNEA), Sp. 

Lejeunea (§ Eulejetjnea) flava, Gottsche, Lindenb., & Nees, Synops. Hepat. p. 373. 
Widely distributed in the Tropics. 

Lejeunea decuesiva, v. d. Sande Lacoste, Syn. Hep. Jav. tab. xiv. 
Our sterile specimen agrees exactly with the figure quoted. 
Distrib. Java. 

Lejeunea (§Miciiolejeunea) ghacillima, Mitt, in Joum. Linn. See. v. p. 115. 
Distrib. Ceylon, Tropical Africa. 

Erullania brunnea, Gottsche, Lindenb., & Nees, Synops. Hepat. p. 441. 
Distrib. S. Africa. 



64 THE PLANTS or MILANJI, NTASA-LAND, 

Metzgeria purcata, Dum. Eec. Obs. JuBg. i. p. 261. 
Widely distributed. 

Metzgeria mtriapoda, Lindb. in Act. Soc. Faun, et El. renii. i. (Monog. Metz. p. 22). 
Distrih. America. 



DESCBIPTIOK OE THE PLATES. 

Plate I. 

Figs. 1-3. Anemone Whyieana, Baker fil. 

Pig. 1. Plantj natural size. 

2. Stamen, enlarged. 

3. Carpel, with style, enlarged. 

Pigs. 4-18. Impatiens shirends, Baker fil. 

Pig. 4. Plant, natural size. 

5. Standard, side view, natural size. 

6. Keel, with spur, side view, natural size. 
7 & 8. Lip, front view, natural size. 

9 & 10. Sepals, natural size. 

11. Stamens, natural size. 

12. Peduncle, with ovary, natural size. 

13. Portion of leaf, showing serration, magnified three diameters, 

Plate II. 

Figs. 1-4. Eriosema shirensis, Baker fil. 

Pig. 1. Standard, magnified two diameters. 

2. One of the wings, magnified two diameters. 

3. Keel, magnified two diameters. 

4. Legume, magnified two diameters. 

Figs. 5-12. Flemingia macrocalyx , Baker fil. 

Fig. 5. Plant, natural size. 

6. Standard, magnified two diameters. 

7. One of the wings, magnified two diameters. 

8. Keel, magnified two diameters. 

9. Stamens, magnified two diameters. 

10. Ovary, with style and stigma, magnified two diameters. 

11. Legume, magnified two diameters. 

12. Transverse section of same to show amount of turgidity, magnified two diameters. 



COLLECTED BY MR. ALEXANDER WHTTE. 05 

Figs. 13-17. Dissotis Johnstoniana , Baker fil. 

Fig. 13. Part of plant, natural size. 

14. The same, in fruit, natural size. 

15. Bud, natural size. 

16. One of tlic small stamens, enlarged. 

17. One of the large stamens, showing the elongate arcuate connective, the figure to the left natural 

size, that to the right being the same enlarged. 

Plate III. 

Figs. 1-6. Choristylis shirensis, Baker fil. 

Fig. 1. Part of plant, natural size. 

2. Portion of inflorescence, natural size. 

3. View of the flower, the two front petals having been removed, enlarged. 

4. Fruit, enlarged. 

5. Stamen, enlarged. 

6. Pistil, enlarged. 

Figs. 7-13. Tryphostemma apetalum, Baker fil. 

Fig. 7. Part of plant, natural size. 

8. Portion of stem, with base of leaf, peduncle, and stipuliform appendage, natural size. 

9. The same, enlarged to show the relative position of the points of attachment to the stem, of the 

petiole, peduncle, and stipuliform appendage. 

a. Stipuliform appendage. 

b. Point of attachment of same. 

c. Point of attachment of peduncle. 

d. Point of attachment of petiole. 

10. Sepal, enlarged. 

11. Inner corona and stamens, enlarged. 

12. Outer corona, enlarged. 

13. Ovary, with styles and stigmas, enlarged. 

Plate IV. 



Figs. 1 & 2. Othonna Whyteana, Britten. 



Fig. 1. Portion of plant, natural size. 
2. Fruit, with pappus. 



Figs. 3-5. Vaccinium afrtcanum, Britten. 



Fig. 3. Branch, natural size. 

4. Flower, enlarged. 

5. Upper part of stamen, with anther, enlarged. 



SECOND SERIES. — BOTANY, VOL. IV. 



66 THE PLANTS OF MILANJI, NYASA-LAND, 

Plate V. 

Figs. 1-6. Erica Johnstoniana, Britten. 

Fig. 1 . Branch, natural size. 

2. Flower, enlarged. 

3. The same, without calyx or corolla_, enlarged. 

4. Upper part of stamea, with anther, enlarged, 

5. Lower side of leaf, enlarged. 

6. Upper side of leaf, enlarged. 

Figs, 7-12. Erica Whyteana^ Britten. 

Fig. 7. Branch, natural size. 

8. Terminal shoot with crowded leaves, natural size. 

9. Flower, enlarged. 

10. The same, without calyx or corolla, enlarged, 

11. Upper part of stamen, with anther, enlarged, 

12. Leaf, enlarged. 

Plate VI. 

Figs. 1-3. Mascarenhasia variegata, Britten & Hendle. 

Fig. 1. Branch, natural size. 
2. Flower, enlarged. 
.3. Portion of corolla- tube, cut open to show stamens, enlarged. 

Figs. 4-6. Stathmostelma reflexum, Britten & Rendle. 

Fig. 4. Branch, natural size. 

5. Flower, enlarged. 

6. Pollen-carrier, enlarged. 

Plate VII. 

Mgs. 1-3, Disa zombaensis, Rendle. 

Fig. 1. Portion of stem, with spike, slightly reduced. 

2. Front view of flower, after removal of odd sepal and lateral petals, slightly enlarged. 

3. Pollinium, enlarged. 

Figs. 4-6. Dlsa hamatopetala, Rendle. 
Fig. 4. Plant, natural size. 

5. Side view of flower, after removal of odd sepal, enlarged. 

6. Lateral petal, enlarged. 



COLLECTED BY MR. ALEXANDER WHYTE. 67 

Plate VIII. 

Figs. 1-5. Vellozia splendens, Rendle. 

Fig. 1. Shoot, slightly reduced. 

2. Flower, natural size. 

3. Leaf, natural size. 

4. Perianth leaf, inside view, showing pair of scales and stamens, enlarged. 

5. Stamen, enlarged. 

Plate IX. 

Figs. 1-5, Thesium Whyteanum, Rendle. 
Fig. 1. Branch, natural size. 

2. Flower, enlarged. 

3. Longitudinal section of same, enlarged. 

4. Fruit, enlarged. 

5. Apex of leaf, enlarged. 

Figs. 6-11. Widdringtonia Whytei, Rendle. 

Fig. 6. Branch, bearing male cones and cluster of fruits, natural size. 

7. Transverse section of a branchlet of W. juniperoides, showing two opposite loaves with single 

resin-canal in each. 

8. The same of W. Whytei, showing the three resin-canals in each leaf. 

9. Piece of branchlet and detached leaf of W. Whytei. 

10. Male cone of same. 

11, Seed of same. 

Plate X. 

Figs. 1-6, Panicum pectinaium, Rendle. 

Fig. 1. Plant, natural size. 

2, 3, 4. First, second, and third empty glumes, enlarged. 

5. Flowering glume, enlarged. 

6. Portion of peduncle bearing capitate hairs. 

Figs. 7-12. Harpechloa altera, Rendle. 

Fig. 7. Plant, natural size. 

8. A spikelet, enlarged. 

9. Flowering glume. 

10. The same, opened, and seen from within. 

11. Pale. 

12. Upper barren flower. 



E . G.Balcer. 



Traks.Likn-, Soc.BoT. SBR.2Vbl.IVPl.l. 




Figs. 1-3. AITEMOJSTE WHYTEANi^, Baker fa. PiqsA^lS. IMPATIENS SHIRENSIS, BcLkerfil. 



E.G.Boiker, 



Teaks. LmN. Soc.Bot. Sbb. 2rVbl.[VPi.2, 




KMorq TJ:!- .dd Bt nth. . 



"WcEit. Jfevon^r. imp. 



riqs.l-'t.ERIOSEMA SHIKENBJS, £c±erM. "Figs. 5-12. FLEMING! A l/tACRO CALYX, BaJccrl-i/.. 



Pigs. 13-1 7. DISS OTIS JOHNS TONIANA, Baker fd- 



E.G.Baker. 



Trans.Linn. Soc.Bot. Ser 2VoI.IV:P1.3. 




RMci-ga.naii!.etlitK. 

Pig6.l-6.CH0RISTYI.IS SHIR^NSTS , 5a>t^/?Z. Pig s .7. 13 .mTPHO S TEMMi^ APE TAI.UM, ^a^ Z-^, 



West ,!NEwmfm tm£. 



BTittein.. 



TRA>rS.LlNW . Soc .BoT . Ser 2 Vol. TVPL .4. 




A '»<\' 



R.MCT-gott del ellith . 




We St . ITe-wmacn. iicna . 



Piqs.l-2.0TH0NNA WHYTEANA, ^-iift/7,. Figs.r^-5 VACCINLUM AFRIGANUM B. 



, tSntten. 



BritteTL . 



Trans Linn Soc Bot . Ser.2.Vo1.TV.P1.5. 






iji fin \S"i 

w C % « iw W/: 11 
'^ ' i '^ ^, ^ i _ 




"B-Morgaix del et litK 




Figs. 1- 6 . ERICA. JOHNSTOmANA , BrOZen. Figs. 7.12 . ERIC^ WHYTEi^ITA, Bria^ 



BTittea & Rendle. 



Trans. LiNW.Soc.BoT .SEH.2VolJV!Pi.6. 




BMoEqan. HbL ct liih. 



"West.lTewttuni-tmji. 

Fig s.l 3 . MASCAKENHASI A VARIE QKi:^, Britten & ife/z^. Piqs 4- 6 . S TATHMOSTELMA PffiPLEXUM, Brat^^ LRendl^. 



I^Tulle 



Trans.Lhw. Soc.Bot. Ser.2:VqI.IV:P1.7. 




■R.MOTg4iL.a«iL.et IMi 



:AENSI>, 



.lAMATOPE TAT , A ^\v..*7^ , 



RenoOfi. 



Traks. Linn- Soc Bot. Ser. 2."Vbl,IVP1.8. 




R.Mjrgeua del.etli-tii 



VELLOZIA 3PLENDEN3, B^ndU. 



"West.'Newmicn.iirrp, 



Bjemcle. 



Trans.Linn. Soc.Bot. SER.2.Vol.IViP19. 




■RMcrrgon. deist lith. - WeatNevanaixirap. 

Figs.l-b.THESIUM IWHYTEANDM , iEa^z^ife. FicjS.G-il.WipDRIITGTOWL-:. WHYT'EI, i?<;/KZfe. 



HaruiLe . 



Tkans . Linn . Soc.BoT . Ser.2 ."VbL rTiPl.lO. 




R .Mcax) «n .dol et lit]i. 



"WestiTIewmari. itirp. 



Fiqs.l-fe.PANICTJM PECTINATUM, iii^y^^. Figs, 7-12. IIABPECHLOA ALTERA, Ay. .<£&. 






LTNNEAN SOCIETY OF LONDON. 



MEMORANDA CONCERNING TRANSACTIONS. 

The First Series of tlie Transactions, containing both Botanical and Zoological contributions, has been completed 
in 30 Vols,, and a few entire seta are still for sale. Only certain single volumes, or parts to complete sets, may 
be obtained at the original prices. The price of the Index to Vols. 1-25 is 8s. to the public, and 6s. to Fellows ; to 
Vols. 26-30, 4s. to the public, and 3s. to Fellows. 

The Second Series of the Transactions is divided into Zoological and Botanical sections. The prices of the 
Botaaical parts of these which have been published are as undermentioned (for the Zoological parts see Zoological 
wrapper) : — 



,r 1 When Price to the 

Volume, TuUkh^l. Public. 

£ s. d. 

Sk^SD SEBIKS. BOTAXY. 

I. Part I. 1^75 S 

F^ .11. 1875 16 

Ba^ HI. 1876 12 

tM% Vf. 1870 lo 

Part ^t. lS7a .... 1 4 O 

Part VX'lSm .... 1 6 

Pint Vn. 1^0* .... 1 4 

^ftTCa, 1880* .... 1 1 

tm IX. 1880. .... 1 

II. Fart-^ 1-1881 ^^ 12 

t&d' m,!"^ 5 

feft 'nt 'Mm, .... 10 

B^MT. IS^v.... 3 

■fefe V* issa..... 3 
■^- It. -mm. .... 13. 6 
■^^^m^-i^ .... 9 6 

^^^^I0^^^\,.. 10 
^^^^■im^,-,..:; 7 

l^^atl%;^^*--.--- «■* '3 4 



Price to 
Fellows. 



Tolume. 



When 
Published. 



... tJ 


(' 


. . . 12 





. , . . 9 





.... 7 





. . . . 18 





.... 1& 


6 


.. .. 18 





. . . . 16 





. . . . 15 





.... 





.... 3 


9 


.... 7 


6 


.... 2 


3 


.... 2 


3 


.... 10 


i> 


.... 7 





.... 7 


6 


5 





.... 2 


6 



Second Skkies.- 
n. Part XI. 1886. 
Part XI L 1880. 
PartXIII. 1887. 
Part XIV. 1887. 
Part XV. 1887. 
PartXVI. 1888. 

III. Part I. 1 ^88. 
Part T I. 1891. 
Part HI. 1801. 
Part IV. 1801. 
Part V. L89L 
Part VI. 1391. 
Part VII. 1892. 
PartVIII. 1803. 
Part IX. 1803. 

Part X. 1804. 

Part XI. 1S04 
IV. Part I. i?^:»4. 



Price to the 
Public. 

£ s. d. 

f.iNi' {conti.ii 



Price to 
Fellows. 



•C'i). 



« 


. . 


. . 


4 


fi 


8 


. . 


. . 'J 


6 





1 7 


. . 


.. I 





i) 


7 


,. 


.. 


5 





10 


u . . 


1) 


1 


r, 


2 


6 , . 


, . 


2 





3 12 


. . 


2 


!4 


<,) 


5 


. . 


. . 


:.i 


J) 


7 


. . 


.. 


5 








.. 


.. 


4 


6 


6 


.. 


.. 


4 


6 


3 


6 .. 


,. 


2 


8 


6 


.. 


. . 


4 


(J 


6 


.. 


. . 


4 


6 


I 8 


.. 


.. 1 


] 





10 


{) . . 


.. 


7 


^; 


3 


4 . . 


. . f> 




i'> 


1 n 


i) 


. , If 


1-" 


n 



r 4* toi 












h:f*X. 
















2nd Ser. BOTANY.] 



[VOL. I¥. PAET 2. 



THE 



TEANSACTIO^S 



OF 



THE LINNEAN SOCIETY OF LOIS^DON. 



ON THE FLORA OF MO [XT KTXABALU 
IN NORTH BORNEO. 



BY 



(y STAFF. Pr.T).. 



( !Utli/l>'/i 



u.fttd Ixj W. T. Tht^eltox Dykr, C.Ar.G., F.E.S , F.L.S.) 




SOLD 



•ANS, SREEF, ANB 



[ 69 ] 



II. On the Flora of Mount Kinahalu, in North Borneo. By 0. Stapf, Bh.J)., 
Assistant for India, Boyal Herbarium, Kew. {Communicated by W. T. Thiselton 
Dyer, C.M.G., F.B.S., F.L.S.) 

(Plates XI.-XX.) 

Eead 15th June, 1893. 

In May, 1892, Dr. G. D. Haviland, then Curator of the Museum at Kuching, Sarawak, 
successfully accomplished an expedition to Mount Kinabalu, in North-east Borneo. A set 
of his botanical collections was sent to Kew, and I was entrusted with the determination 
of the plants and the description of the new species. In examining them I soon found out 
that our set contained not only all those species of Sir Hugh Low's collection from the 
same mountain, which were described by Sir Joseph Hooker in the ' Icones Plantarum ' 
(1852), and in the ' Transactions of the Linnean Society ' (1860) — with the exception of 
Nejpenthes, of which there is no specimen in Dr. Haviland's collection — but also nearly 
all the other species from the same collector wliicb were preserved, unnamed or 
provisionally named by Sir Joseph Hooker, in the Kew Herbarium. The same was the 
case with a set of specimens collected by Mr. F. W. Burbidge. I decided to include 
them with Dr. Haviland's collection, particularly as they supplemented it in a very 
useful way as specimens, as well as by the notes which Sir Hugh Low had attached to 
them, and not less because I felt that I must not silently pass over the prominent 
share which Sir Hugh Low had in the exploration of that remarkable mountain. Thus 
the result was a complete enumeration of the plants hitherto collected on Mt. Kina- 
balu. The indications of altitude and of many other details on Sir Hugh Low's and 
Dr. Haviland's labels and numerous notes and sketches in Mr. Burbidge's diary, which 
he kindly lent me, in combination with the descriptions of the general features of 
the mountain contained in Low's report on his first ascent (Journ. Ind. Arch. vi. 
1852, pp. 1-17), in Spencer St. John's article " Observations on the North-west Coast 
of Borneo " (Journ. B. Geogr. Soc. xxxii. 1862, p. 217-233), and in the same author's 
book ' Life in the Eorest of the Par East,' in Burbidge's ' The Gardens of the Sun,' 
and in Whitehead's * Exploration of Mount Kinabalu,' have encouraged me to draw a 
sketch of the general differentiation of the vegetation of the mountain in z ones of altitude 
and in botanical formations.* The presence of Australian and Antarctic types in the 
flora of Kinabalu was already known from Sir Joseph Hooker's publications and from 
occasional remarks in the * Genera Plantarum ' and in Bentham's ' Plora Australiensis,' 
but it is much more pronounced in the collection now under consideration. This fact, 
and the very remarkable phytogeographical relationships of Kinabalu, induced me to 
also add a paragraph on the phytogeographical affinities of the flora of the mountain. I 

* " Pflauzenformationen " of Grisebach, " formations vegetales," "formazioni vegetali." 
SECOND SERIES. — BOTANY, VOL. IV. ^ 



70 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

am well aware that conclusions drawn from about 360 species, perhaps one fourth or 
one fifth of the phanerogamic vegetation of the mountain, cannot he considered final, hut 
they may yet mark a considerahle progress if we keep in view that all the hitherto 
published species from Kinabalu amounted to less than 20 species. In working out 
the collection I met with the same difficulty which is alluded to by almost every botanist 
who has had to deal with the Malayan flora — the difficulty which arises from Blumc's and 
particularly from Miquel's practice of describing species from very incomplete material, 
and from the fact that not a few of the determinations in the collections which were 
distributed from Leyden are untrustworthy. However, by revising the Malayan material 
of those genera with which I had to deal in the Kew Herbarium, I have succeeded 
generally in bringing together trustworthy types of most of the species hitherto described, 
or at least of those which possibly might come under consideration, thus laying a solid 
foundation. 

The Orchids have been determined and described by Mr. Ridley, Superintendent of the 
Botanic Garden in Singapore; the Perns by Mr. J. G. Baker, Keeper of the Royal 
Herbarium, Kew, and the Mosses by Mr. C. H.Wright, Assistant in the Royal Herbarium, 
Kew, and by Mr. W. Mitten; whilst Mr. C. B. Clarke has been kind enough to name 
most Cyperacese and to place his descriptions of two new species of Cladium at my 
disposal. It very fortunately happened that Dr. Haviland arrived at Kew just before 
this paper was ready for press. He obliged me deeply by going through the general part 
of the manuscript with me and by assisting me with his valuable experience gained on 
the spot. This entailed, of course, various corrections and modifications, adding, 
however, I hope, to the value of the paper. He allowed me also to insert an abstract of 
an account of his expedition which was originally drawn up for private use. 

I here avail myself of the opportunity to express my most sincere gratitude to 
these gentlemen for their kind help, and to add particularly my heartiest thanks to 
Mr. W. T. Thiselton Dyer, the Director, for the permission to work out this most 
valuable collection, and to publish the results, and to Professor D. Oliver and my 
colleague Mr. W. Betting Hemsley, for their assistance in many difficult cases. 



I. HISTOEICAL ACCOrNT OF THE BOTANICAL EXPLOEATION OF MOUNT KINABALU. 

Eeom Sir Hugh Low's First Expedition (1851) to 
Mr. Whitehead's Exploration (1887, 1888). 

Mount Kinabalu was first visited by Mr. (now Sir) Hugh Low, who nearly 
reached the summit, in 1851. A collection made on this occasion was sent to 
Kew, and the more striking species were described by Sir Joseph Hooker in the 
'Icones Plantarum,' vol. ix. (1852), whilst an account of the ascent was published 
by Sir Hugh Low himself in 'Journ. Ind. Arch.' vi. (1852) pp. 1-17. In 1856 
Mr. Thomas Lobb, the well-known collector, attempted to reach the mountain, but 
he was forced by the natives to return. Some of Lobb's plants in the Kew Herbarium 
marked " Borneo " may have been collected on this abortive expedition, but there is no 




DE. O. STAFF OX THE FLOKA OF MOUNT KINABALU. 71 

evidence for it, except in the case of a species of Finanga which was ohtained at Bongol, on 
the Lower Tampassuk. In 1858 Sir Hugh Low, accompanied by Mr. Spencer St. John, 
Consul-General for Borneo, twice repeated his attenij^t to explore Kinabahi. The first 
time, in April and May, Sir Hugh Low was comjielled to stay at Kiau, on the foot of 
the mountain, whilst Mr. St. John ascended it almost to the top. The second time, in 
July, he was more successful, and they both reached a peak which is only second in 
height to the summit. It was then that the splendid Pitcher-23lants were collected, which 
Sir Joseph Hooker described and figured in the ' Transactions of the Linnean Society,' 
vol. xxii. (1859), and made the mountain at once famous. A very exhaustive 
report on both expeditions may be found in Spencer St. John's ' Life in the Far East.' 
Several later attempts to ascend Kinabalu had no botanic importance, l)ut in 1877 
Mr. F. W. Burbidgc went there to collect for Messrs. Veitch of Chelsea. A set of 
dried plants gathered on Kinabalu was sent to Kew, together with others from Labuan, 
from the Lawas River, and other places in North-cast Borneo. As, however, the labels 
in many cases do not indicate the locality beyond the general printed heading " North 
Borneo," I have omitted those without the special indication, unless it was evident from 
Burbidge's book or diary that they were from Kinabalu. In 1887 and 1888 Mr. John 
Whitehead, an ornithologist, undertook several expeditions to Mount Kinal)alu. In 
1888 he stayed for more than four weeks at an elevation of about 8000 feet, and he 
ascended the top under very favourable conditions. Mr. Whitehead collected a few 
plants, which he presented to the Natural History Museum, South Kensington. His 
splendidly illustrated book 'Exploration of Mount Kinabalu' contains not only a 
detailed account of his expedition with several very fine illustrations representing the 
mountain as seen from different sides and elevations, but also numerous remarks 
which throw light on the general character of the vegetation. The last and, from 
a botanical point of view, most successful expedition was Dr. G. D. Haviland's, 
in March and April 1892, on the results of which the greatest part of the present 
paper is based. As Dr. Haviland has not published an account of his expedition, I 
asked his permission to insert an abstract of an account drawn up by him for private 
use. This abstract, which follows, consists of the copied passages of the account 
which I thought of interest in connection with the subject of this paper, and 
Dr. Haviland kindly undertook the revision of it. 

Dr. Haviland's Expedition (1892). 

Dr. Haviland writes : — *' We left Kuching on March 1st, and reached Gaya late on the 
evening of the 5th. Here we stayed till the 10th, whilst our packages and provisions were 
being carried by boat to Tawaran. We reached Tawaran early on the morning of the 11th. 
The river is small, and there is a bar at the mouth. The Government offices are situated 
about an hour's pull up the river. Beyond there is a large open plain, free from jungle, 
and thickly populated by Dusuns and their buffalos. The paddi harvest was over, the 
ground very dry, and the river low. On the 12th we sent our packages up river to 
a house at the east end of the plain, near the foot of the hills, whilst we and the coolies 
walked on foot, and we reached the place long before the boat. The river was about 

l2 



72 DE. 0. STAPr ON THE FLOEA OF MOUNT KINABALU. 

30 yards broad, but in places scarcely more than 2 feet deep, the whole way across with 
a sandy bed. At daybreak on the 13th the land-breeze from Kinabalu was very 
strong, and the thermometer read 21° C. "We kept along the north bank of the Tawaran 
till we came to a tributary, which we crossed and followed a short distance before 
beginning the ascent of the spur between it and the main river. The highest point we 
reached was 1800 ft. above the sea; from this we could see the Tawaran curving round 
us to the south. A little further on we reached a scattered village, where we passed the 
night. At daybreak on the 14th the temperature was about 23° C, although our 
altitude was 1500 ft. ; a thick mist filled the valley below us, which must have been the 
head of the tributary we crossed yesterday. We got off early, and still kept along the 
mountain-ridge in a direct course for Kinabalu. Everything was very dry, though in 
the afternoon we saw rain at a distance in the direction of Kinabalu. About midday we 
reached our highest point, 2700 ft., whence the ridge turns northward, but we kept our 
eastward course, descended into the valley, and crossed a stream running southwards 
into the Tawaran ; its bed had an altitude of 1500 ft. "We still kept eastward and 
before long came upon the Tawaran itself, now a rapid stream with a very stony bed, 
and about to our knees in depth ; its altitude was about 1500 ft. Keeping along it and 
crossing it, we soon reached the village of Bongol; here the Tawaran bends round 
coming from the southward, a tributary only coming from the eastward. The village 
presented quite a different appearance to any in Sarawak territory, for it stood on 
pasture-land closely grazed by buffalos. Pinang-palms and cocoanut- trees were scattered 
about in clumps, and climbing up them w^ere strong tall plants of sirih [Piper Betle). 
We got off early on the morning of the 15th, and after following the river a short 
distance, again turned eastwards and ascended the single ridge here separating the 
Tawaran and Tampassuk. The altitude of our highest point was about 3100 ft. ; from 
thence our descent to the Tampassuk was very steep. This river was much larger than 
the Tawaran at Bongol, but its altitude above the sea almost the same, and we forded it with 
the water scarcely over our knees. "We kept along its eastern bank till we reached the 
village of Koung, situated on a flat pasture-land mth hills rising abruptly from its 
sides and the Tampassuk coursing through it. A drizzling rain at 3 p.m. indicated our 
proximity to the mountain and a change of weather. 

" We crossed the river on the morning of the 16th, and kept along its western bank. 
On reaching the Kadamaian, the most southerly branch of the Tampassuk, our party 
divided, the larger crossing it and ascending the northern slopes of the ridge between the 
Kadamaian and Penokok, getting splendid views of Kinabalu up the Penokok valley ; 
but I with some Dyaks kept to the valley of the Kadamaian ; the water was low, and we 
crossed and recrossed it several times. At an altitude of about 2000 ft. we left the river- 
bed and ascended the southern side of the ridge between Kadamaian and Penokok, 
reaching the village of Kiau, which is situated at about 3000 ft. on the southern slopes 
of this ridge. 

" The next day we went out to look for plants and birds. I descended northwards 
towards the Penokok. The top of the ridge was about 3500 ft. above the sea, and 
thence we got a splendid view of the south-western precipices of Kinabalu from their 



DR. 0. STAFF ON THE FLOEA OF MOUNT KINABALF. 73 

base. Rocky irregular spurs and hollows covered with old jungle formed the heads of 
the Penokok and Dahombang, and descended to an altitude of about 8000 ft., where 
these rivers entered cultivated land, and running westAvard, separated only by a strij) of 
land, do not join till they reach the entrance of the narrow valley, into wliich the 
Kinitaki from the north and the Kadamaian from the south also flow% and which is the 
commencement of the Tampassuk. The Haiahaia is a tributary of the Penokok, which 
joins it just as it enters the cultivated area ; both seem to drain from the great southern 
spur of Kinabalu and its western offshoot, on which thc^ village of Kiau stands. The 
Dahombang is larger than the Penokok, and seems to drain from a part of the very base 
of the great precipices. Looking across this comparatively broad Penokok-Dahombang 
valley, we saw it bounded on the other side by a steep ridge somew^hat higher than tlie 
one we were on and covered by old jungle. Beyond this w^is the Kinitaki, a still larger 
river, draining the greater part of tlie huge precipices, and running parallel to the 
Dahombang towards Mt. Saduk Saduk till turned southwards by it, and entering tlie 
Dahombang a little below its junction with the Penokok. Beyond the Kinitaki must 
be a watershed joining Mt. Saduk Saduk with the Maripari spur of Kinabalu, and 
beyond this must be the rivers mentioned by the Dusuns : Lambun, Pengataran ♦, Liban, 
and Wariu, entering the Tampassuk in its lower course. The Kadamaian comes from 
the very top of Kinabalu. Its head at Low-'s Gully runs down the southern slopes of 
Kinabalu, passing between the heads of the Kinitaki, Dahombang, and the Penokok 
on the west side, and a branch of the Sugut on the east side, then curving westwards it 
passes south of Kiau, then north-west till it joins the Dahombang just below that river's 
junction with the Kinitaki, thus forming the commencement of the Tampassuk in a 
narrow valley on the south-west side of Mt. Saduk Saduk. In Mr. St. John's map 
Kalupis is used for what the Kiau men told us was called Kadamaian. On being asked 
they said Kalupis was a small branch of the Kadamaian. 

" Next morning v,g were up at 4.30 a.m. in hopes of starting early, but it was midday 
before we were off. The Dusuns had other matters on hand, and they took us only half 
a mile to a kladi-fieldf in the valley below. 

" Next morning, the 19th, Dusuns came from the village to carry our things, and we 
started up stream. On oiu- left was the granite peak of Kinabalu, just overtopping 
the Kiau ridge, and in front was a narrow valley of old jungle, headed by a ridge some 
6000 ft. in height, being the long southern spur of Kinabalu, and along this ridge lay 
our way to the mountain-top. We soon left the kladi-fields, and taking to the river-bed 
at an altitude of 2600 ft. we passed into the jungled vaUey, whose sides w^re at first 
quite precipitous. We had much clambering over huge slippery boulders, till by 
midday we reached an altitude of 3200 ft. The northern side of the valley w^as still 
inaccessibly steep, but the southern w^as no longer so, and on starting again we gradually 
ascended this, still keeping on our eastward course. We now caught sight of a 
mao'nificent cascade, the Kadamaian waters entering the valley by pouring over the 
ridge in front. The stream seemed scarcely to have cut into the rock at all, and for the 
first 1500 ft. was nearly perpendicular. At 2 p.m. we came to a great overhanging rock. 

* Mentioned as " Pantaran " by Wlutehead. f Cdladium escuUntum. 



74 DE. O. STAFF ON THE FLOE A OF MOUNT KINABALU. 

The place was 5000 ft. above the sea, with a good stream of water close by. The Dusuns 
called it Lohang ; here we stayed a week. The rock was large enough to shelter all our 
party, but, being convex vertically, the shelter was not good. In heavy rain the water 
ran down almost to the base of the rock before it dripped off. Pacing to the west we 
never saw the sun ; shut in by mossy trees we scarcely saw the sky. At daybreak the 
thermometer was 14° C. ; daily the mists set in about 11 a.m., and the rain began at 
midday. 

" Many Bornean heights form knife-like ridges, a few feet broad at the top. Along 
these is a track probably kept open by deer, on either side of which grow plants. These 
are almost always in flower and love the sunshine, which can be here felt even through 
the mist. Among these plants are notably Vaccinium, Rhododendron, and sometimes 
Pitcher-plants. A few feet lower the vegetation is completely changed, but nowhere 
have I seen this change more marked than on this ridge above Lobang, which was the 
southern spur of Kinabalu. The sides shelved steeply down, and on them grew the 
sub-summit shrubs with lanky bent and angled stems, densely clothed with long damp 
moss and lichen, and gy:owmg fro7ii the slope rather than upwards in the struggle for the 
.light. Beneath them were no other shrubs or herbs but tangled moss-covered roots. 
Along this zone, which was more than 100 feet in height, we usually found a fallen fruit 
closely resembling a chestnut. Below this zone is common jungle of mixed trees, shrubs, 
and herbs. No bird or squirrel was the same as in the low country, and there were 
no monkeys. At an altitude of about 6000 feet was a tree, from whose hollow a bear 
had recently removed a nest of bees. On the morning of the 22nd we followed a well- 
marked path which we found led to Temburungo, the place where Mr. Whitehead formed 
his camp at an altitude of 7750 feet. On reaching the top of this ridge we found in 
front of us the Sugut, behind us the valley of the Kadamaian from which we had just 
come ; on our left was an upper valley of the Kadamaian above the waterfall, a much 
larger valley than I had expected to find and full of tall jungle- trees. In its lower 
part it was not steep ; above this it stretched suddenly upwards and northwards to the 
southern slopes of the granite cap of Kinabalu. We had to follow the ridge between 
this vaUey and the Sugut a little way before reaching Temburungo. This was just such 
a swampy spot as in England is loved by Sphagnum, Carices, Heaths, Gentians, Asphodel, 
and Brosera. The Drosera, however, I could not find ; the Asphodel had its represen- 
tative ; the Gentian was a dwarf. There was no running water, but pools of it in holes 
which had been dug by Mr. Whitehead. We had sunshine and splendid views to the south 
and to the west. I followed the ridge further to 8000 feet, when it became obscure. On 
the 26th we went up to Temburungo. Kinabalu had been visible up till noon, and after 
that was hidden only by white clouds. Up till now the sun had made the place quite 
warm, but at 3 p.m. the rain began and the temperature rapidly fell to 13°*6 C. 

" On the 27th I started for Pakapaka. We kept along the big southern spur of Kina- 
balu, having sources of the Sugut on our right and the upper Kadamaian on our left. 
The rise is as great as from Lobang to Temburungo ; but it is longer and more gradual. 
We were soon in mist, and, as all the jungle-plants were dripping with moisture, got wet 
through. We reached Pakapaka about midday. The cave was formed by a single 



DR. O. STAFF ON THE FLOE A OF MOUNT KINABALU. 75 

slab of granite, and though small it promised good protection, and the mountain 
torrent just at its mouth gave a good supply of Avater. The mist and rain increased. 
Everything got damper and damper. The boulders of rock over whicli I had at first 
walked and climbed after plants were gradually submerged in a foaming roaring torrent, 
descending the hill at an angle of about 20°. Some small shrubs grew behind the 
large boulders, being protected by them. These were so encompassed and shaken by 
the water that the torrent could seldom have risen much higher than it was. All 
the afternoon the cold wind blew up the stream into our cave. At 5 p.m. the 
temperature of the air was 11° C, of the water in the torrent 10° C. The aneroid 
read 20*84 ; the boiling-point was by one thermometer 291" F., by another 90 C. 
There were quantities of flowering shrubs round our cave and down both sides of the 
torrent. 

" On the 28tli the temperature at daybreak was 8 C. About 7 o'clock wc started, and 
were soon again on the ridge overlooking the Sugut. The ridge broadened out, and at 
about 11,000 feet we found ourselves in quite a different region at the foot of the granite 
cap, which stood steep and almost bare somewhat on our left. Along our ridge the 
gradually dwarfing shrubs ran up yet another 1000 feet, forming a thick jungle, and 
reached to the eastern end of the granite cap. "Wc avoided tliis jungle, and keeping on 
the Kadamaian side passed over a flatter area, where many little streams from the granite 
cap collect and unite before beginning the rapid descent past Pakapaka cave. Ilerc 
were patches of shrubs in flow^er, patches of bare rock, and patches of mossy swamps, 
wliere grew buttercups, Potentillas, and a small white Gentian. It was a delightfully 
warm place when the sun was out. We soon began the ascent of the rough granite cap, 
sloping at first but getting gradually steeper, till we were obliged to make use of ledges. 
A little higher up the steepness disappeared, and we found ourselves at the foot of a long 
inclined plane of granite, sloping to us from the north. Low's Gully was in front, but 
we kept to the west and reached the top of the sharp northern ridge about its middle. 
It overlooked a lofty precipice, which, with the nearly precipitous sides of a rugged ridge 
running northwards to Mt. Nohan on the east and a buttress of the largest north-west 
peak on the west, formed the boundary of a deep valley with perpendicular sides and a 
narrow outlet to the north-west in the direction of the Pengataran. 

" The plants on the top were stunted, only a few inches high, growing only in the 
crevices of the rock, to which they had a very firm hold, so that it was difficult even with 
both hands to collect them with their roots. In two hollows shrubs were growing. 
One was at the foot of Low's Gully, where was the blood-red Rhododendron : the other 
higher up near the foot of St. John's Peak ; here I collected Eiibiis Loicii. We stayed 
about a couple of hours on the top collecting all the time, and returned that afternoon 
to Temburungo. It was the last fine day we had, and the mist prevented us from ever 
again getting to 12,000 feet. On April the 2nd we again moved to Pakapaka, and 
stayed there till the 9th : every day the clouds and mists came up about 9 a.m. and rain 
began about noon, and by 1 o'clock the stream in front of our cave had become a roaring, 
foaming, impassable torrent, which subsided in the night. 

" At 8 A.M. on the 8th the thermometer read 9°'5 C, the aneroid 20"84 the boiling-point 



76 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

was 194°-07 P. The altitude of tlie Pakapaka care must therefore be 2700 ft. above 
Temburungo, or 10,450 ft. above the sea. 

*' On the 10th at 3.30 p.m. the temperature at Temburungo was 15"^-^ C, the aneroid 
22"90 inches, the boiling-point hj one thermometer 92°'8 C, by another 198°"7 P. The 
altitude of Temburungo must therefore be 7750 ft. above the sea. 

" On the 11th we returned to Lobang, and on the 12th to Kiau. On the 14th we moved 
to the bed of the Penokok river, where we camped till the 24th. On the 18th I visited 
Maripari, where the aneroid about noon stood at 24*6 inches, which gives the altitude at 
about 5000 ft. Maripari seems to be on the spur of the great north-western peak which 
runs out to Mt. Saduk Saduk, separating the Kinitaki from the headwaters of the 
Pengataran. On the 24th we commenced our return to the coast. We found the rivers 
swollen and unf ordable, and that we bad fallen on weather unusually wet even for Kina- 
balu." 

II. GEOGRAPHY, GEOLOGY, A^"D METEOEOLOGY. 

Orogeaphy and Geology. 

Mt. Kinabalu is situated in Keppel Province, British North Borneo, 20 miles from 

Ambok Bay and 30 miles from Gaya Bay. It is the highest mountain of the 

Malay Archipelago, reaching an altitude of 13,698 feet according to Capt. Belcher, 

and is not volcanic like most other high mountains of this region. Mt. Kanabalu 

proper is a huge mass of eruptive (?) granite, which emerges with precipitous, almost 

everywhere inaccessible, sides from a much and deeply dissected highland. The 

summit itself is not much moj^e than 2 miles long ; but it sends a long spur out from 

near its eastern end, running N.N.E. This spur culminates in a long, jagged, and 

narrow ridge, between 10,000 and 11,000 feet high, and breaking off in equally inaccessible 

precipices ; it is most probably formed also of granite. Another spur runs from the 

south-eastern part to S.S.W. This spur, or rather range, however, has a very different 

character ; it begins where it parts from the granite cap at an altitude of about 

11,000 feet, descending soon below 8000 feet. St. John estimates its length at 20 miles ; 

but it is probably much longer and perhaps in direct orographic connection with the 

mountains from which the Padas and the Kimanis rivers drain ; yet it does not seem to 

rise anywhere to an altitude of more than 6000 to 7000 feet, whilst it is known to descend 

in the Tambunan Pass, 30 miles south-west of Kinabalu, to 4700 feet. This range, the 

main range of the highland of Kinabalu, consists, for several miles south from the 

granite cap, " at first of sandstone, then of black shale, almost as hard as rock " 

(Spencer St. John, Journ. E,. Geogr. Soc. xxxii. p. 233). The ridge is narrow, but, like 

the steep slopes, covered with dense forest. The jagged northern spur suddenly breaks 

off, leaving a deep gap between itself and a high peak of conical shape, the altitude of 

which is estimated at about 8000 feet. Otherwise there is no continuation of the hi^h- 

land in this direction. Numerous other spurs shoot out from the flanks of the granite 

cap and the northern spur almost in every direction, leaving narrow and steep vallevs 

between them, but not one attains any considerable length. Those on the western side of 



DE. 0. STAFF OX THE FLOKA OF MOUNT KIXABALU. 77 

the mountain branch off at an altitude of 5000 to 0000 feet, rajiidly descondini* to tlie 
bed of the Tampassuk and to an altitude of 1000 to 2000 feet, and thus havini; enormous 
precipices, 8000 and moz*e feet high, above them. Tlie south-western spurs branch off at 
a higher elevation ; yet they are by no means longer. A few steep conical peaks of 
moderate height, as for instance the Saduk Saduk, near Koung, rise from these spurs, 
west and north-west and at no great distance from Kinal)alu. All these spurs seem to 
consist of sandstone in the lower parts. On some of them a rock, designated as 
serpentine, has been found cropping out near the granite cap, and AVhitehead mentions 
also '* shale." They are evidently all of the same geognostic character as the main 
range. From the east the lowland of the basin of the Sugut Kiver approaches very 
closely to Mt. Kinabalu. 

As to the geological age of the sandstone and of the Schistose formation of the highland 
we have not much more than suggestions to make, based upon the analogy with other 
parts of Borneo which are better known. But Posewitz (' Borneo,' p. 197 &c.) is probably 
right in attributing the same geological character to this highland whicli distinguishes 
the greater part of highland Borneo and the tin islands Banca and Billiton, which are 
particularly well known. If this be so, the schistose rock would represent wliat is called 
the Old Malayan Schistose I'ormation, wiiich would tlierefore form the backbone of the 
highland, from which the granite of Kinabalu proper emerges as an isolated nmss, the 
product of a local but powerful eruption. The sandstone, on the other liand, which 
skirts the spurs and of which the w'hole or the gretiter part of the hilly country between 
Kinabalu and the north coast is composed, would be Tertiary. The Old Malayan Schistose 
Formation was hitherto generally considered as Palaeozoic, particularly Devonian, and 
Posewitz places even the eruptions of granite and diorite which are not unfrequent in 
this region in the Devonian period ; but Martin pointed out a few years ago that the 
Schistose formation is much younger, perhaps even Cretaceous. Ilowever this may be, 
it does not materially affect the assumption that the highland of Kinabalu is old land. 
According to Posewitz, Borneo was an extensive archipelago in the early Tertiary period, 
consisting of several large and numerous small islands, and among the larger ones was the 
highland of Kinabalu, with Kinabalu itself at its northern end. To judge from the rate 
at whicli denudation takes place at present the island must have rcciched a very consider- 
able altitude even then, and the mountains may have been still loftier. The fact that the 
highland of Kinabalu is not only old land, but old highland, is very important with 
respect to the history of its flora. 

I quote below a few passages from Whitehead's book which, in combination with 
Dr. Haviland's account on pp. 71-76, will give the reader a better idea of the physiognomy 
of this remarkable mountain and the surrounding landscape. The first passage, taken 
from pp. 58 and 59, refers to the view from near the coast, north-west of Kinabalu : — 

" The country inland is much broken up by sandstone hills of low elevation, generally 
bare of trees, and in this month (February) parched and brown, covered with one of the 
tropical curses, lalang grass. At the back of these lalang-covered hills are others topped 
with forest, sometimes sharp-peaked and rugged ; at the back of all, towering above every- 
thing, is that superbly grand mountain Kinabalu. This wonderful mountain looks 

SECOND SERIES. — BOTANY, VOL. IV. M 



78 DE. O. STAFF OJs" THE TLOEA OF MOUNT KINABALU. 

more imposing than many mountains of a much greater altitude, from the fact that it is 
an almost isolated mass, at the highest end mounting up to the height of 13,698 feet 
(Eelcher) and finishing off abruptly in sheer precipices of 10,000 feet ; its proximity to 
the sea-coast, being less than forty miles in a direct line, allows of it being viewed at once 
from the lowest level, none of its imposing grandeur being lost by a series of intervening 
plateaus which dwarf so many mountains." 

The following passage is taken from page 101, and refers to the view obtained from 
Melangkap, a village north-west of Kinabalu and close to it : — 

"At a glance one could see that the mountain on this side was a sheer precipice and 
quite inaccessible, the water falling thousands of feet without touching the rock. . . . 
Eetween the main buttress (the summit) and the long jagged ridge (the northern spur) is 
a huge gorge surrounded on three sides by sheer precipices of bare rock, the entrance to 
this gorge being from the western side, facing Melangkap. At the most southern 
extremity the mountain suddenly terminates at its greatest altitude by precipices 
of rock. The top along the whole distance is bare of vegetation and very rough ; here 
and there great scars, formed by rock-slips, may be noticed. Though Kinabalu is only a 
few miles distant, the intervening country is a mass of forest-clad ridges, which look from 
here very steep, and through which the Pantaran (a tributary of the Tampassuk) and 
other streams have cut, in the course of ages, deep channels. . . . The village of 
Melangkap is situated at the end of a huge spur which runs round in a slight curve 
almost to the base of the great buttress of Kinabalu. At the Melangkap end the 
altitude is over 1000 feet, but this spur gradually rises on nearing the mountain to over 
5000 feet. It is covered with a dense forest-growth and branches off in several directions, 
the Pantaran flowing along its base, the head-waters of this river falling from the top of 
the buttress itself. On looking to your left, when facing the mountain (from the village 

of Melangkap), another lofty spur sweeps round from the other end of the mountain. 
This spur is more broken up and about 4000 feet in altitude. . . . Directly in front of 
Melangkap, and between the two mentioned ridges, the country is a mass of huge spurs, 

mounting up, as Kinabalu is neared, to 5000-6000 feet." 
I may add that the first of the two spurs mentioned parts from the granite cap very 

near Maripari, a locality spoken of by Dr. Haviland in his account. This will help the 

reader to join Whitehead's w ith Haviland's description, which refers to the south-western 

spurs and to the southern main range. 

Meteorology. 
The ' Handbook of British North Borneo ' contains very valuable information on the 
meteorological conditions of British North Borneo, derived mainly from observations at 
Sandarakan and other places on the coast. We may assume that the same conditions 
prevail also on the coast w est and north-w^est of Kinabalu and for some distance inland. 
But they undergo a considerable change as we approach the very foot of the mountain, 
and still more as we ascend it. There is, according to the ' Handbook,' a " true wet season," 
extending over the months November, December, and January, and often also parts of 
October and Pebruary, and a '* true dry season," following almost immediately and 



DE. 0. STAFF ON THE FLORA OF 3I0UNT KIXAliALU. 19 

lasting to May, whilst the rest of the year divides between what is called a " moderate 
wet " and a " moderate dry season." The " true dry season " seems, however, to set in 
rather earlier, at the end of January or early in Eebruary, on the coast near Kinabahi ; 
and February has the reputation of being the driest month even on Kinabalu itself. 
Yet, whilst the two main seasons are well marked on the coast and drought of considerable 
duration (as much as 22 days) is known during the dry season, when the open country 
occasionally assumes a parched appearance, the climate is very equable all the year round 
on Kinabalu. Even the driest month is here still very wet, and days on which no rain 
falls are very rare. Thus extreme humidity and equability arc the most prominent features 
of the climate of Kinabalu. The daily periodicity of the liydrometeors (/. e. hydrometcoric 
phenomena) which is so characteristic for the high mountains of Java, and, in fact, more or 
less for all the liigh mountains within the Tropics, is very pronounced on Kinabalu. The 
nights and the mornings are generally clear, with the exception of the bottoms of the deep 
valleys, where mist collects in the early hours. At 9 or 10 o'clock clouds form on the 
mountain, wliich by midday obscure it entirely down to 5000 or COOO feet. Then rain 
— sometimes also hail or sleet — falls almost incessantly till late in the evening, when the 
air clears again. There is, however, a remarkable difference in the extent of the belt of 
clouds or mist. In Java the clouds form generally batween 5000 or 0000 and 8000 feet, 
leaving the summit more or less free and comparatively dry . This is not the case on 
Kinabalu, where this belt reaches up to between 10,000 and 11,000 feet, and where also 
the highest part of the mountain seems to receive an enormous amount of rain, the waters 
of which partly fall in cascades from the top ridge and partly rush as a foaming, roaring 
torrent past the Pakapaka cave (at an elevation of almost 10,500 feet). The annual 
amount of rainfall on the coast ranges from 100 to 157 inches, or from that of Singapore 
to not quite that of JBuitenzorg; but on Kinabalu itself it must b3 enormously increased. 

The monthly maxima of temperature on the coast range from 80'*88 F. in Eebruary 
1885 to 92 '65 in April of the same year, and the monthly minima from 71 '46 in 
December 1883 to 78^*0 in April of the same year. 

Spencer St. John, who was twice on Kinabalu, in May and in July 1858, found that 
the thermometer at Kiau, on the south-west foot (at 3000 feet), never marked above 77° 
during the day, and varied from 66 to 69 during the night. On the Maripari spur at 
4700 feet, it marked in July 76 (mean) at midday, 60° at 6 a.m., and 56° (mean) at 6 p.m. 
At the Pakapaka cave (10,450 feet) he found the temperature ranging from 36°-5 to 43° 
during three nights in May and 41°'125 (mean) at night in July. On the summit at 
1 P.M., exposed to mist and rain, the thermometer marked 52°. Whitehead made daily 
observations at Temburungo, on the southern main range (7750 feet), from the 1st of 
February to the 3rd of March. He found the temperature very uniform, the nights 
varying between 42 and 52°, the days between 60° and 70°. There are various other 
readings of the thermometer made occasionally by travellers, and in different months, all 
of which come within the range indicated by the figures quoted. When St. John observed 
the thermometer fall to 36° all bushes and trees were " fringed with hoarfrost." This 
makes it probable that even frosts occur occasionally on the top of the mountain ; snow, 
however, has never been seen on Kinabalu. 

M 2 



80 DE. 0. STAFF O^^ THE FLOEA OF MOUNT KINABALU. 

Anotlier peculiarity of the climate is the small amount of insolation which the slopes 
and the bottoms of the valleys and even the ridges receive. After three hours' sunshine 
they get only dispersed light, and not very much of that, as a very great part of the light 
must be absorbed by the thick belt of mist hovering above or resting upon them. This 
state of things is naturally intensified on the western, north-western, and south-western 
sides of the mountain, where the sun rises too late over the huge granite mass to 
shine long on the deep and narrow valleys, and some of the hillsides exposed to the 
north certainly never receive even an hour's sunshine. 

III. GEXEEAL CHAEACTEE OF THE VEGETATION. 

A. Zones of Altitude and Botanical Eormations. 

Relying on the reports of the travellers, and the dried material before me, I distinguish 
roughly the following zones of altitude on Kinabalu : — 

1. The zone of the plains and low hills , or briefly the hill zone, from the literal zone 
which skirts the coast as a belt of varying l^readth up to 3000 feet. 

2. The lower mountain zone, from 3000 to 6000 feet. 

3. The upper mountain zone, from 6000 to 10,500 feet. 

4. The summit zone, from 10,500 feet to the very summit (13,698 feet). 

1. The mil Zone (up to 3000 feet). 

I have taken up in the present paper only the flora of that part of the hill zone 
which forms the very foot of Kinabalu, beginning at the village of Koung, at an 
altitude of 1600 feet, with the exception of a few plants from the Ulu (upper) Tawaran. 

This is the inhabited part of Kinabalu, almost entirely occupied by cultivated land 
and secondary forest (" young jungle ") which springs up rapidly from the clearings as 
soon as they are abandoned. Towards the coast, part of the plains and hillsides are 
covered with lalang-lalang, but this does not reach the foot of the mountain. The 
primary forest (" old jungle ") is hardly at all represented within the limits of this zone. 
The secondary forest is, like the primary forest, essentially evergreen, and has, of course, 
all the characteristics of a true tropical forest. 

The descriptions contained in the travellers' reports do not attempt any philosophical 
classification of the vegetation from a physiognomical and biological standpoint, and 
they are too vague and incomplete to allow the reader to form more than a very general 
idea of the differentiation of the vegetation, nor can the dried material and the notes 
attached fill this gap. I must limit myself therefore — here and later on — to a few 
remarks concerning those principal formations which I find distinctly recognizable. 
These are, in the hill zone : — 

a. The Secondary Evergreen Tropical Forest [*' Young Jungle "]. — I find 40 
phanerogams and 7 vascular cryptogams referable to this formation. Three of the pha- 
nerogams are trees, 21 shrubs, 7 climbers, and 8 herbs. Not one of these is mentioned 
anywhere as particularly striking or forming a prominent feature in the physiognomy 
of this forest, if we except two species of JBauhinia and perhaps Musscenda coccinea, which 
attract the traveller's attention by their extremely showy flowers, and a few large ferns. 
Yet there must be here, as anywhere else, a number of species which are more pro- 



DE. 0. STAFF OX THE FLOKA OF MOUNT KIN ABA LU. 81 

minent or more numerous in individuals than otliors, tlieroby bocominp^ loadinj^ features 
and giving the forest a specific physiognomical and biological character which varies 
according to its composition. In fact, travellers mention certahi plants and cert^iin 
types of vegetation, as climbers, epiphytes, palms, &c., as especially characteristic, and, 
moreover, I am told that the natives know very well how to distinguish not only young 
and old jungle, but also the various types of both. But collectors do not seem to 
have made an attempt to gather specimens methodically of these more prominent 
and commoner types in order to ascertain what is their systematic position. This 
may be largely due to the difficulty of getting such specimens and to want of time ; 
but also to the prevalent illusion that the commoner forms to which the eye soon gets 
used are less desirable in collections, and that the discovery of new species is the sole 
aim of the explorer. Yet what could be more important in exploring a new country 
— from a scientific as well as from a practical standpoint — than to grasp, above all, 
the broad salient features, and to exhibit them iu descriptions and in authentic 
specimens ? I here mention this, not in order to cast any blame on the explorers of 
Kinabalu, but because I have very often felt, in studying parts of the Malayan flora, 
how useful and desirable it Avould be if travellers and colh^ctors would pay more 
attention to this most important branch of phytogeography — I should call it descriptive 
phytogeography — which found so skilful an interpreter for the flora of Java in 
Junghuhn, many years ago. 

To return to the " young jungle " formation of the hill zone, I may mention plants not 
represented in the collection, such as various palms referred to the genera Areca^ Pinanga, 
Calamus by Burbidge, and a species of Korthalaia which I recognized from one of 
Burbidge's sketches. Bamboos play also a prominent part in this forest, and they are 
said to rise in immense feathery clumps to 50 or GO feet on the river between Koung 
and Kiau, and to skirt the torrents and streamlets of the higher valleys with luxuriant 
growth. How far the two species of Bambuseae, represented in the collection, share 
in forming these clumps and belts I do not know. In clearing the land the " old jungle " 
is never so completely destroyed as not to leave now and then solitary trees or groups of 
trees, and even larger plots of the original vegetation, and it seems that the presence of 
tree-ferns on the hillsides near Koung and elsewhere must especially be traced back to 
such an origin. 

As to the systematic character of this vegetation, so far as it is represented in the 
collections, see the table on p. 119. 

b. Cultivated Land. — The cultivated land is limited to the regions below 3000 feet. It 
consists of clearings made in the jimgle by felling and subsequently burning the wood. 
The most important crops are kaladi or kladi (Caladium esculentum) smd rice; then 
follow sweet-potatoes, yams, bananas, tobacco, gourds, melons, cucumbers, chillies — i. e. 
several varieties of Capsicum fruticosum and Capsicum annuwn, Piper JBetle, &c. Of 
palms, the Sago and Cocoa palm, species of Areca, and Oncospernunn filamentosum are 
now and then seen near the houses of the natives, and occasionally also orange- and lime- 
trees and a kind of Artocarpus. There are a few weeds, gathered in the kladi-fields or 
round the houses, in the collections ; they are the commonest Malayan species, and no 
particular interest attaches to them. 



82 DE. O. STAPr ON THE FLOEA OF MOUNT KINABALU. 

2. The Lower Mountain Zone (3000 to 6000 feet). 

This zone is equivalent to Junghuhn's " Zweite Gewachszonc " (second botanical zone) 
or " Gemassigte Kegion " (temperate region), which he, however, limits to from 2000 to 
4500 feet. 

According to St. John, a " fine jungle " occupies the slopes from 3000 to 6000 feet on 
the " main sjour " (the southern main range), whilst it is partly replaced by an exceedingly 
stunted vegetation as far down as 4500 feet on the Maripari spur, probably, as he 
suggests, on account of the nature of the soil, " being formed of decomposed serpentine, 
containing much peroxide of iron " ; but above that there is again fairly-sized forest 
even in exposed situations. A similar forest occupies, it would seem from Whitehead's 
book, also the north-western spur on the Pantaran. We may infer that this forest 
is occasionally interrupted by other formations or more or less modified : for instance, 
along the torrents and streamlets, on exposed ridges, or in connection with the varying 
character of the soil and the exposures and inclination of the slopes. That this is the 
case may also be understood from occasional remarks in the reports of travellers. 
These reports are, however, too incomplete to allow me to give more than a very 
general description of the principal formation, which no doubt predominates almost 
exclusively, and a few notes on another formation of very limited extent, but very 
pronounced character. These formations are : — 

a. Frimary Evergreen Forest [" Old Jungle," " Schattenreiche Hochwaldung " 
(umbrageous high-forest) of Junghuhn]. — I refer to this formation about 145 phanero- 
gams, 17 vascular cryptogams, 17 mosses and liverworts. Most of them were collected 
on the path from the Kadamaian above Kiau to the ridge of the main range, either 
along the river or on the slope of the main range facing west or north-west. The 
rest were gathered in the valleys of the Penokok, Haiahaia, Dahombang, and Kinitaki, 
and chiefly from the sides of the torrents, or on the Maripari spur, which was 
ascended from the south-west. There is no doubt that the border vegetation is 
comparatively better represented than that of the interior of the forest, which is 
generally very difficult to penetrate. This primary forest is described as abounding 
in fine tall trees, creepers and epiphytes, commonly also in undershrubs, and the 
ground as carpeted with ferns and mosses, which increase in quantity and luxuriance 
with the elevation, enveloping trunks and boughs. Herbs intermingle with shrubs, and 
are associated with ferns and mosses. Fifteen species are designated trees by collectors, 
but 11 of them expressly " small trees." Of the remaining four, Ilitrephora Maingayi, 
Sterculia translucens, and F>ysoxylon cauliflorum may be mentioned. A few slender, 
graceful pahns {Areca''^) are also met with throughout this zone, but there are no 
specimens in the Herl^arium. The shrubs are represented by 63 species, but 30 of them are 
indicated only from the upper limit of the zone (5500-6000 feet), and some of these reach 
a much higher level, being-' probably true ridge-plants which occasionally descend lower. 
Twelve shrubs are liubiacece, 8 EricacecB (including Vacciniew), 7 Myrsmacecc, 6 Melasto- 
macetB, 4 EivpJwrhiacece, 3 Myrtacece, 3 Anonacece^ and 3 ^tyracece. It must, however, be 
noted that the majority of FricacecBi MelastomacecBy and Styraoece, and the 3 Myrtacem 



DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 83 

were not collected below 5500 feet. This sequence is not materially altered if we add 
the few woody climbers which are in the collection : 2 Jfelasto77iaccop, StrchJosa urticina^ 
Melodorum khiabaluense, yEschynanthus magnifica, and Smilax hcterophylla^ the last 
two from above 5500 feet. Rattans are also repeatedly mentioned as very frequent, 
though no specimen has been collected. Nepenthes Bajah and Nepenthes Edwardsiana 
are common on the Maripari spur, whilst they do not seem to descend into the lower 
mountain zone on the main range, owing to their greater need of light, which confines 
them more or less to the ridge, which is below^ GOOO feet on tlic ]Maripari spur, but above 
6000 on the main range. Bamboos are another characteristic feature, growing in thick 
clumps beneath the trees, as on the main range, or skirting open places on tlic INIaripari 
spur, or rambling and creeping and " smothering everything," as on the Pantaran spur ; 
but none of these bamboos have been collected. Of the herbaceous plants about 55 
species may be referred to tliis formation, amongst them a dozen orchids, many of 
these herbs growing chiefly along the torrents, as Begonia and Lnpatiens, and only a few 
being restricted to the ui)per limit. Tw^elvc are Orchidete, 10 Mclastomacecey 7 Urticacecs, 
5 Buhiacece, 5 Begoniacece, and 4 Gesneracea;. Amongst the rest Viola serpens^ Impatiens 
pUdyphylla, and Bhlomis rugosa may be specially mentioned. The type of epiphytes is 
represented by Rhododendrons, which, like otlier Ericaceae, have a particular tendency to 
assume this mode of life, and by numerous orchids, ferns, and mosses ; but I have no 
means of defining them more precisely. The ferns fi*om this zone number 16, of which 
9 are limited to the uppermost part, mostly extendmg into the next upper zone. This is 
particularly the case with Hymenophyllum dilatatum and 5 species of TrichoinaneSj 
plants dependent on excessive moisture, and Gleichenia circinata. The 17 mosses and 
liverworts from this zone w ere all collected at or above 5000 feet, except a Campylopus, 
which ranges from 3200 to 9000 feet. 

A rather different kuid of primary forest is mentioned by Whitelicad from high al)Ove 
the village of Kapar ; it was distinguished by the absence of thick undergrowth, caused, 
as he says, by the shade of the high trees. 

b. Bogs. — The presence of boggy places on the Maripari spur is indicated by a small 
number of • plants which, in a similar association, are found on the boggy spot called 
Temburungo, on the main range, but at a much higher elevation. They arc Drosera 
spcithulata, JJtricularia orbicidata, Patersonia Bowii, Aletris foliolosa, Enocaulon 
Sookerianum, and three sedges — Cladiitm borneense, C. samoensey and Schcenus melano- 
stachyus. 

3. Tipper fountain Zone (6000 to 10,500 feet). 

This zone corresponds roughly with Junghuhn's "Vierte Gewachszone" (fourth 
botanical zone) or " Kalte Region" (cold region). When defining the lower mountain 
zone I pointed out that some of the most characteristic elements of that zone are limited 
to the uppermost part, from 5500-6000 feet, and that not a few of them extend above 
6000 feet. This, taken in itself, would suggest hardly more than the presence of a belt 
of transition where the elements of the conterminate zones mingle. I have, however. 



84 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU, 

reason to believe tliat we shall in future be able, with a better knowledge of the 
mountain and its vegetation, to distinguish an intermediate zone analogous to Junghuhn's 
third zone or cool region. This zone will probably comprise the uppermost part of my 
second zone (upward from 5500 feet) and the dense and high, hitherto entirely unexplored, 
forest which occupies the broad and deep depression from which the cataract of the 
Kadamaian falls (see p. 74) and the surrounding slopes upward to near the ridges. It 
would ascend in the sheltered parts of the valley probably to 9000 feet, but leave out 
the ridges from about 7000 or 6000 feet or still lower down. If this should actually be 
the case, the name '' upper mountain zone " would more appropriately be applied to this 
liypothetical zone, whilst the ridges would form a zone by themselves of considerable 
longitudinal extent, but very narrow when measured across, for this ridge-vegetation 
is less determined by the decrease of temperature than by the amount of light and the 
exposure to wind. 

The diflSculties of cutting a path through the primeval forest hitherto compelled all 
travellers who visited Kinabalu to keep to the very ridge, between 7000 and 10,500 feet, 
whence our knowledge of the vegetation of my third zone is almost exclusively limited 
to that of the ridge of the main range. Hence also the want of precision in the definition 
of what I call tlie lower and upper mountain zones. 

Dr. Haviland (see p. 74) mentions a " sub-summit " vegetation forming a narrow 
belt of more than 100 feet high, immediately below the proper ridge. He describes it 
as very different from the " common jungle " below as well as from the ridge-vegetation. 
It evidently constitutes a very marked formation; but in the absence of a sufiicient 
representation of this " sub-summit vegetation " in the collection I must refrain from 
discussing it, and limit myself to the ridge-vegetation, which is represented very well 
indeed in the collection, namely by about 110 phanerogams, 24 vascular cryptogams, 
and 3 mosses. There are two formations distinctly recognizable, JBrimary Evergreen 
Dwarf Forest and Bogs. 

a. Primary Evergreen Dwarf Forest [Dwarf " Old Jungle "]. — It practically occupies 
the whole ridge so far as it is known, with the exception of a few open places. This forest 
consists of small trees and tall shrubs ranging from 10-20 feet, and varying in denseness 
from almost impenetrable thickets to open spots. The trees are stunted, twisted, and 
weather-beaten, often being bent across the path. The trunks and branches are clothed 
inches deep with dripping moss and festooned with long beard-like lichens. Only conifers 
grow into fine trees here in some more favourable places. The tendency of some of the 
trees and shrubs to grow gregariously is distinctly noticeable. The foliage of these trees 
and shrubs is often crowded on short and thick branches ; the leaves, sessile or supported 
by short and stout petioles, are very coriaceous, of a dark green colour, glabrous — at 
least above — and glossy. They show a marked tendency towards oval and round 
forms, particularly near the base, and the generally entire margins are not seldom 
recurved. 

Eleven species are designated trees, but, as already mentioned, only a few conifers 
attain under existing conditions a considerable height, namely Fodocarpus cupressina, 
the scale-leaved form of which has been mistaken by some travellers for Casuarina, 



/ 



BR. O. STAPP ON THE ELOEA OE MOUNT KINABALU. 85 

and Dacrydium elatum, both very common. Jlyrtus flavida and Rhododendron cunei- 
folium assume a yew-like appearance, whilst j^a-oups of llhododendron ericoides exhibit 
the Erica type, and numerous dwarf oaks [Quercus Uavilandii) are representative of a 
closely-allied form (Q. pruinosa var. alpina, Jungh.), found under similar conditions on 
the high mountains of Java. The shrubs blossom nearly the whole year, and many of 
them very freely. Not less than 9 species of Rhododendron adorn this ridge with their 
brilliant llowers collected in dense clusters or scattered among the dark foliage. They, 
in addition to 5 species of Vaccinium and 7 species of JDip lycos ia, render the order of 
Ericacece (inch Vacciniece) the most conspicuous and numerous of the phanerogamic 
vegetation of this zone. Some of them become occasionally epiphytal here as elsewhere. 
Then follow Mubiaceoi with 10 species, two of which are climbers, whilst one (Psychotria 
densifolia) was found epiphytal on trees. Pcav Ruhiacem, however, reach 8000 feet, and 
only one [Uedyoils macrostegia) goes beyond 9000 feet. Myrtacece number 6 species, 
amongst which Leplospernmm recunnim is most conspicuous, covered all over with 
innumerable white llowers like snow. Myrsinece are represented by 5 species, among 
them being the dwarf Emhelia minutifolia and the widely-spread Myrsine capitata. 
Ternstroemiacece and RosacecB number 4 species. Among the former Schima hrevifoUa 
represents the Camellia type, whilst 3 of the 4 Rosaccie tare species of Riilnis, and 
Stranva^sia integrifolia replaces Photinia integrifoUa, which is found in Java under 
similar conditions. Urticaceoi are represented by three shrubby figs, all l)elow 9000 feet, 
and CupulifercB by 1 Quercus and 2 Castanopsis, the latter also limited to the lower part 
of the zone. Styracea number also 3 species {Symplocos). Among the remainder I 
may mention particularly EIcbo carpus sericea, a representative of the Javan Acronodia 
punctata, Blume ; Rhus borneensis, a very singular member of the section Venenatae ; 
the large-flowered Rolyosma Hookeri ; an endemic species of Leucopogon ; 2 Laurinece ; 
and 2 species of Loranthus having large flower-heads and living chiefly on Rhodo- 
dendrons. Among the climbers may be especially mentioned Smilax Icevis and a 
very common Calamus, the latter upward to 9000^ feet. The Pitcher-plants, however, 
are by far the most peculiar feature, though most of them are only locally frequent. 
There are not less than 5 species found on this ridge, some climbing in the trees, as 
i^epenthes Loiciiy others rambling in the shrubs or straggling over the ground. 

The herbaceous vegetation, so far as it forms a distinct part of the forest, is repre- 
sented by 15 species, a remarkably small number, of which 8 are Orchidem^ 3 Gesneracece, 
and one a common Malayan grass. To these 2 or 3 herbs might be added, which 
though distinctly foreign, namely Austral-Antarctic types, seem to be more or less 
intimately associated with this formation. They are Nertera depressa, Euphrasia 
borneensis, and Trachymene saniculcefolia. Balanophora elongata, a common parasite 
living upon the roots of EHcacece, is here as frequent as it is in Junghuhn's fourth 
zone or " cold region." 

The fern vegetation attains here its most luxuriant development, and, as it seems, the 
o-reatest variety, being represented in the collection by not less than 21 species. Two 
are tree-ferns, and one, Cyathea Uavilandii, was found at 10,500 feet. The delicate 
Trichomanes and a species of Hymenophyllum are limited to the lowest part, whilst the 

SECOND SERIES. — BOTANY, VOL. IV. N 



86 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

more robust species of Folypodium were chiefly collected at the upper limit. Lycopodia 
become more frequent, and the wiry trailing stems and branches of L. casuarinoides 
— according to Dr. Haviland — often cut the hands of the traveller who tries to penetrate 
the thickets. 

The epiphytic vegetation is certainly very luxuriant also here ; but besides the shrubs 
already mentioned as growing occasionally epiphytic on trees only three orchids are 
expressly designated epiphytes, and ferns, mosses, and lichens in a very general way. 

In comparing this description of the ridge forest on Kinabalu with Junghuhn's 
description of the primeval forest of the " cold region " of the high mountains of Java, 
the reader will note several facts which are the more significant as the number of species 
taken into consideration by Junghuhn and by me is about the same (112 species from 
Kinabalu against 90 to 100 by Junghuhn). The first thing which impresses itself upon 
the reader is the evidently great general physiognomical resemblance — see, for instance, 
pp. 427, 428, and 449 of Junghuhn's ' Java, seine Gestalt, Pflanzendecke, und innere 
Bauart,' vol. i. (2nd edition) ; the next, that the species and even the genera are mostly 
not the same, but mutually representative. On the other hand, the presence of rattans 
up to 9000 feet; of so many Hhododendrons, also above 7500 feet; of Podocarpus 
cupressina, one of the most characteristic elements of Junghuhn's third zone, upward to 
11,000 feet ; the prevalence of ferns, the absence of subarboreous Coniposita3 (Anten- 
naria), and the scantiness of the herbaceous vegetation — if we may rely on the collectors 
in this point — constitute diiferences worthy of consideration. 

b. Bogs. — These are confined to a few very limited places, where, on account of 
particular conditions of the soil, water collects and trees and shrubs cannot get a footing. 
The vegetation is materially the same as in similar places on the Maripari spur (see 
p. 83), but a dwaif Genticma and Trachymene saniculafolia, an Australian type, are 
very remarkable additions. 

4. Summit Zone (10,500-13,698 feet). 

There is no exact equivalent amongst Junghuhn's zones, as the nature of the top 
portion of Kinabalu and of the summits of the Javan volcanos is so entirely diff'erent ; 
but the few analogies which we may expect lie with certain parts of Junghuhn's 
fourth zone. 

Beyond the Pakapaka cave (10,450 feet) the forest dwarfs down to a mere shrubbery, 
continuing along the ridge towards the eastern end of the granite cap, but otherwise 
broken up in scattered patches, ascending to 12,000 feet in the narrow and steep gully 
which bears Sir Hugh Low's name and affords the only means of reaching the 
summit. Above that shrubs were found only in two hollows. On a flatter area, where 
many little streams from the granite cap collect and unite before beginning the rapid 
descent past Pakapaka, boggy patches with buttercups, potentillas, and gentians are 
scattered between patches of shrubs and patches of rocks, whilst the remainder of the 
scanty vegetation clings to the rocks, satisfied with the little soil collecting in crevices 
and holes. By far the greatest part of this zone, however, is occupied by bare rock. 



DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 87 

Dr. Haviland took particular care to collect specimens of as many species as ho could 
find here, and we may therefore consider this flora fairly completely represented in the 
collection. It comprises 52 Phanerogams, 4 Vascular Cryptogams (species of Lyco- 
podium), and 5 Mosses, and they are distributed over three formations, namely, evergreen 
dwarf bush, bog, and the vegetation of the rocks. 

a. Evergreen DioarfBush. — This formation, consisting of a narrow strip extending along 
the uppermost part of the main I'ange, and of smaller or larger isolated patches, partakes 
of the character of the dwarf forest into which it passes gradually below 11,000 feet in 
many ways, being primarily the product of the extreme development of those conditions 
which influence the ridge-vegetation so effectively and of the scantiness of the soil which 
the denuded granite — nowhere very favourable to plant-life — affords. The shortening 
of the axial portions and the consequent crowding of the foliage are carried to an 
extreme. The general character of the leaves is the same as in the forest immedLately 
below; but the size is mostly much reduced, and the colour still more intense, imbuing 
the dark green of the vegetation with a brownish tint. The number of species repre- 
sented in this bush is still very great if we consider the limited area and the high 
elevation. The shrubs number 26 species, of which 5 are restricted to the very lowest 
part, and another 5 were not collected above 11,500 feet, whilst still 8 species were 
found above Low's gully, at 13,000 feet. The list of shrubs is headed also here by 
EricaeecB with 5 species, then follow Mi/rsinea with 3 species ; but they do not extend 
above Low's Gully, nor do the 2 corifers of this zone, namely, Fhyllocladus hypophjllus 
and Podocarpiis nervlfolla, var. hrevifoUa. The other orders, each of 2 shrubby 
species, are TernstroemiacedB, Rosacece, Ilyrtacece, and Bubiacece. The 8 species 
gathered above Low's Gully, at 13,000 feet, are Miihus Loicii (in a dwarf form, only a 
few inches high), Leptospermum recurmtm, and Leucopogon siiaveolensy all known also 
from the third zone ; then Eurya reticulata, a shrub characteristic also of the highest 
regions of Sumatra and Java, and Stranvcesia iniegrifolia, both descending below 
12,000 feet, and Symplocos huxifolia, Coprosma Hookerl, and Drapetes ericoides from 
12,000 to 13,000 feet. Drapetes ericoides forms low and very compact tufts, — hardly 
deserving the designation of shrubs — of a very peculiar habit which is not represented 
otherwise in the flora of Kinabalu, and it is altogether a foreign element. It grows in 
the crevices of rocks, and might be perhaps better included with the rock-vegetation. 
There does not seem to be much herbaceous vegetation immediately associated with the 
bush, as nearly aU the herbs from this zone are expressly stated to grow either in open 
boggy places or on the rocks. The few exceptions are Eria grandis, very common from 
10,000 to 12,000 feet on the ground and in the shade of the bush, Bulbophyllmn montense, 
an epiphyte, Balanophora elongata, and perhaps some of the sedges. 

The fern- vegetation is entirely replaced by Lycopodia, which, associated with mosses, 
cover the ground, particularly along the edges of the bush. 

b. Bogs. — Places exhibiting a more or less boggy character seem to be more frequent 
here than in the lower zones, particularly in the flatter portion, immediately at the 
foot of the granite slope; yet they nowhere cover a considerable area. The flora 
of these bogs consists of not a dozen phanerogams, of little interest so far as their 

n2 



88 DE. 0. STAPF ON THE FLOEA OF MOUNT KINABALU. 

biological characters are concerned, but very important on account of their geographical 
relations. On these I shall dwell later on. It will be sufficient to mention here the 
generic names in order to give an idea of the composition of this peculiar vegetation. 
They are JRmmnculus, JPotentilla, Saloragis, Gentiana, Havilandia, a new genus of 
Boraginea? allied to Trigonotis and 3I?/osofis, probably also Trachrjmene and JEuplirasia^ 
Aletris, and a few sedges. 

c. Vegetation of the Rocks. — There is probably little justilication to speak of the 
vegetation of the rocks as truly distinct from that of the bogs, as the conditions of 
plant-life afforded by the small accumulations of soil in crevices cannot — in this 
particular case — materially differ from those of the bogs. The few plants which I 
have to enumerate here are a variety of Potentilla leuconota, Haloragis micrantJm, 
Trachymene smiiculcefolia, in a very stunted form, Pilea JoJmiana, Tlatyclinis stachyoides, 
Schcenus apogon, and two grasses limited to the very top, namely, varieties of Deschampsia 
jleocuosa and of Agrostis canina. 

B. Ox SOME Biological Eeatures. 

Under this heading I wish to deal briefly with some of the more obvious biological 
characters of the vegetation, so far as they may be ascertained from the dry material 
and the notes of collectors. 

1. Foliage of the Forest Vegetation. — The woody vegetation is almost exclusively 
evergreen, and I doubt whether there are more than a few shrubs or trees in the Kinabalu 
collections having periodically deciduous foliage, though I find about 25 species dis- 
tinguished by thin membranaceous leaves and about as many by leaves which, although 
much firmer, still might be called membranaceous or papyraceous. It is very remarkable 
that the species having thin membranaceous leaves are, perhaps with a single exception, 
limited to the lower mountain zone, and there chiefly to the part below 4000 feet. The 
other class, to which species of Saurauja, Vrophyllum, Lasianthus, Ilelastoma, &c. 
belong, are more equally distributed. In any case the woody plants having truly coria- 
ceous leaves amount in all the zones taken together to almost 80 per cent., and to more 
than 80 per cent, above 6000 feet. These coriaceous leaves (see also p. 84) are generally 
glabrous, and where there is an indumentum in a young state it quickly disappears from 
the upper side, at least after the unfolding of the lamina, usually leaving a glossy 
surface of a deep green. The lower side is oftener clothed with a distinct indu- 
mentum, which is either softly tomentose or silky, or consists of stiff spreading hairs. A 
tomentose covering occurs, for instance, in Folyosma hracteolata, Geimsia farinosa, and 
in Quercus Savilandii, plants of widely different systematic affinities, habit, and 
habitat. Elceocarpus sericea, Hubus lineatus, and JJrophyllum lineatum have silky 
leaves, the latter chiefly along the nerves which are unprotected in the bud. The 
leaves of five of the Myrtacece also are silky, though only so in a very early stage and on 
the middle nerve alone, evidently a protective contrivance. A strigillose indumentum is 
very distinct in the young state of Melastoma, Marumia, Bissochceta, and Saurauja, yet it 
disappears gradually when the leaves are fully grown, whilst it is persistent, for instance, 



DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 89 

in Diplycosia chrysotJmx and D. rufa. These coarse, often bristly and short hairs are 
here, as in other cases, much more prominent on the young branches, where they usually 
persist, after the leaves have becouie glabrous, on one or on both sides. 

Another very characteristic feature is the almost perfect absence of deeply divided 
or compound leaves. Excepting Muhus — Dysoxylon caulijloruniy Emdia tenuistylat 
Guioa pleiiropteriSi Fithecolohium blgeminumy and Artlirophyllum dlversifolium are 
the only species having compound leaves. This tendency to simple leaves becomes 
particularly prominent in the single-leaved Hhus borneen^is, which is very nearly allied to 
Mhus sticcedanea^ a species having 2-8 pairs of leaflets. 

2. Flowers. — The comparative scarcity of showy flowers even in the most luxuriant 
parts of the Tropics has repeatedly been alluded to by travellers. It finds an em2)hatic 
expression in the Kinabalu collections, although we must not forget that the tree- 
vegetation, the tall climbers, and the epiphytes of the two lower zones arc very incom- 
pletely represented in the collection, mainly on account of the difficulty of gathering 
specimens from them, whilst they are all within easy reach above, where the vegetation 
assumes a stunted character. On the other hand, we may expect that plants having 
relatively conspicuous flowers are more likely to be collected than those which are less 
attractive. This has been the case in the hill zone in quite a prominent degree. "With 
reference to those species which have either very large and gaily-coloured flowers (e. g. 
Melastoma, Bhododendron), or where smaller flowers are gathered in large and consjiicuous 
inflorescences (e. g. Ixora, Musscenda), or where they are scattered in abundance over 
the surface (e. g. Leptospermum recurvum)^ if we indicate them as Class J., those having 
very small and inconspicuous flowers (e. g. Microtropis, Vitis, EuphorUacecB, Vrticacece) 
as Class III., and the remainder as Class 11., and if we exclude grasses and sedges, we 
arrive at the following result : — 

Class I. 

Hill zone 22 per cent. 

Lower Mountain zone . 8 
Upper Mountain zone. 12 
Summit zone 12 

The figures derived for the hiU zone are of little importance, as this zone is very 
poorly represented in the collection, and the endeavour of collectors to gather just 
a few of the most conspicuous plants, as Bauhinia or Musscenda, is quite manifest. As 
to the lower mountain zone the reader will remember that the Hhododendrons which 
are foremost among Class I., become prominent only near the upper limit, and the same 
applies to three or four other plants ranging in Class I., thus leaving very few really 
showy plants in the lower and middle part of the second zone. If we further take into 
consideration that none of the more conspicuously flowering plants of this zone grow 
gregariously, and that the herbs having gaily-coloured flowers (e. g. Melastomaceae, 
Impatiens) are scattered amongst ferns and moss and beneath shrubs, we must admit 
that the observations made by travellers as to the want of showiness in tropical vegetation 
are wholly borne out by the data derived from the collections from Kinabalu so far as the 



Class IL 


Class III. 


58 per cent. 


20 per cent. 


62 


30 


54 


34 


52 


36 



90 BE. 0. STAFF OjN^ THE FLOEA OF MOUNT KINABALU. 

lower mountain zone is concerned. The case lies, however, differently with the ridge- 
and the summit-vegetation. The relative number of sj)ecies with very showy flowers is 
here about twice as large as in the greater part of the lower zone, and the effect must be 
the more striking where the number of individuals increases at the same time, and above 
all on account of the stunted growth of the forest, the fl.ower- treasures of which are placed 
immediately before the traveller's eyes. The proportion of the three classes is the same 
in the summit zone as in the ridge flora, but the vegetation is altogether too scanty to 
l)e impressive, and particularly the herbaceous vegetation of the bogs and rocks is far 
from displaying that gaudy show of colours which we are used to associate in our mind 
with the idea of a truly alpine flora. 

"White (or white tinged with yellow or pink) and red are the prevalent colours if we 
exclude the flowers of Class III. More than 90 per cent, of the flowers of the lower 
mountain zone of which the colour was ascertainable are either white or red, and the 
proportion of these two colours is 2 : 1 in this zone, whilst yellow and still more blue 
or violet are very rare. The ridge and the summit vegetation has 75-80 per cent, white 
or red flowers, the proportion of both colours being again 2 : 1, whilst the rest is yellow. 

3. Fruits. — Eleshy fruits, mostly berries, are remarkably frequent. They amount to 
35-40 per cent., whilst the percentage of plants with seeds, apparently adapted by their 
smallness for dispersion by wind, is 25-28 per cent., whether we take the vegetation 
above 3000 feet as a whole or zone by zone. A few plants, as Parameria, Soya, and 
JEschy^iantlms, possess seeds with long tufts of hairs constituting a sort of flying con- 
trivance, and the pappus of some of the Compositaj acts in the same way. The latter, 
however, are weeds, confined to the hill zone. Otherwise there is very little that is 
remarkable in the fruits and seeds with regard to contrivances which might be con- 
sidered to be means of dispersion. The large percentage of fleshy fruits and minute seeds 
would seem to point in an eminent degree to the interference of animals and currents of 
air in their distribution. But the fact that a fruit is fleshy and therefore, in our opinion, 
attractive to certain birds is no proof that it is really devoured by them, and still less 
that it is dispersed by them, particularly over a wide area. Observations on the spot, 
examinations of the contents of the crop and the intestines of shot birds, collections of 
fruits and seeds found in them, and experiments to ascertain whether the seeds retain 
their germinating power after having passed the digestive organs of the animals, are the 
only means, worthy of science, for elucidating the part which the presence of a " fleshy " 
structure in fruits and seeds plays with regard to distribution. The rest is mere 
conjecture. We possess already a few observations of this kind from the Malay 
Archipelago, but they are too few to generalize from them. The distribution of some 
of the berry-producing plants (as, for instance, JRubus roscefollus, Nertera dej^ressa, or 
Dianella ens'tfolla) l)y birds, either in the present or in some more or less remote time, is 
highly probable from the area over which these species range. Yet the fact that so 
many of the endemic species possess fleshy fruits is extremely unfavourable to the 
assumption that the general object of this particular structure is to act, in the first i)lace, 
as a means of dispersion, especially over wider areas. It may, of course, occasionally be 
the case, and with particular efi'ect within a very narrow area, or at a very slow rate and 



DE. O. STAPP ON THE FLOKA OF MOUNT KINABALU. 91 

in combination with other agencies. The same reluctance will be necessary in drawing 
conclusions from the minuteness of seeds and fruits apparently adapted to the transport 
by wind. Here, again, very many of the Kinabalu plants having such seeds or fruits are 
endemic, or they live in sheltered, quiet places, where the action of wind must necessarily 
be very limited. 

IV. ENDEMISM. 

To speak of the endemism of a district so little known and forming part of a likewise 
imperfectly explored flora is a very difficult task, and I do it only under a certain 
reservation. Better material, and particularly more experience concerning the variability 
of the plants, will no doubt enable us in future to reduce some of the endemic forms to 
already described species of a wider range, and others may be found in other parts of 
Borneo or of the Archipelago. I have, however, reason to believe that the number of 
endemic species from the three upper zones will not be very considerably lessened, 
especially if we take the highland of Kinabalu, not exactly Kinabalu proper, as basis. 
I have, for instiince, lately worked up the whole Kew material of Thalami flora* of North 
Borneo, from the west frontier of Sarawak to the east coast of British North Borneo, 
and besides gone through Dr. Haviland's more recent and very ample collections from 
Sarawak, without coming across more than one of the species which I had originally 
taken to be endemic on Kinabalu. Thus the fact that the flora of Kinabalu is extremely 
rich in endemic species may be considered as well established. 

The endemic Phanerogams number 199 out of 3i2, or 58 per cent., but their distribution 
over the four zones is as follows : — 

Zone. ... I. n. III. IV. 

Number and percentage 1 8 (19 per cent.) 89 (57 per cent.) 74 (65 per cent.) 30 (59 per cent.) 

of endemic species. J ^ ^ ' ^ ^ ^ ^-^ ^ ^ ' 

Among the 8 endemic species of the hill zone are three or four species which some 
botanists, holding a wider view with regard to the limitation of species (for instance in 
the sense of Bentham in his ' Elora Australiensis '), would perhaps refer to already known 
species of a wider range. This would reduce the percentage of endemic species within 
the hill zone to not much more than 10 per cent., a comparatively very low figure. 
Above 3000 feet the percentage of endemic species is very high. It is a little higher 
in the ridge- vegeiation than below 6000 feet, but the difference is not very great. In 
the summit zone it seems again to decrease. We must, however, distinguish here 
between the bush and the vegetation of the open land (bogs and rocks), the proportional 
share of endemic species in the bush being 2 in 3, but only 1 in 2 in the open land 
veo-etation. Thus it is manifest that the decrease of endemism in the summit zone 
is solely due to the flora of the open land, which is about as rich in species as the bush, 
and contains more elements of a wide range, particularly amongst sedges and grasses. 
The bush itself, however, exhibits the same extraordinary endemism which distinguishes 
the forest of the main range. 



92 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

If we arrange the more important orders represented above 3000 feet accordmg to 
their proportional share of endemism, we get the following sequence : — 

Orders. Total. Endemic. Per cent, 

Ruhiacea 30 29 96-6 

EricacecB 32 29 90-6 

MyrtacecB 9 8 88-8 

Urticace<B 13 11 84'7 

StyracetB 6 5 83*5 

Orchidea 24 18 75-0 

Gesneracea 7 5 71*4 

Melastomaceee 21 14 QQ'Q 

TernstroemiacefB 6 4 66*6 

Anonacece 5 8 60*0 

Nepenthaceae 5 3 60*0 

MyrsinacecB 17 9 53*0 

Euphorbiacea 6 3 50*0 

miices 49 11 22-4 

There axe 17 phanerogamic genera with more than 3 species above 3000 feet ; they 
are arranged according to their proportional share of endemic species : — 

Genera. Total. Endemic. 

Urophyllum 5 5 

Psychotria 4 4 

Lasianthus 4 4 

Vaccinium 7 7 (?) 

Elatostemma 5 5 

Diplycosia 11 10 

Rhododendron 12 10 

Symplocos 6 5 

Bulbophyllum 6 5 

Embelia , 5 4 

Begonia 6 4 

Sonerila 5 3 

Nepenthes 5 3 

Eria 4 2 

Ardisia 7 3 

Rubus 5 1 

Carex , 4 

Thus it appears that Diplycosia, Rhododendron, and Vaccinimn have produced the 
greatest number of species peculiar to Kinabalu ; Ericaceae are therefore the great 
leading feature in the flora of Kinabalu. 

The generic endemism is surprisingly small, being confined to the genera ffavilandia, 
a Boraginea allied to Myosotis and Trigonotis, and Scyphostegia, a somewhat doubtful 
Monimiacea and a perfectly isolated type. 



DE. 0. STAFF ON THE FLOKA OF MOUNT KlNABALl. 93 

V. AFFINITIES. 

I place four tables at the end of the general part of my paper, one for each zone, 
sho-vving the distribution of the species enumerated in the special part, witli the exee|)- 
tion of the commonest weeds of the hill zone, in order to facilitate tlie review of the 
complex geograj)hical relationships of the flora of Kinabalu. The reader will find some 
explanatory notes on the areas and the signs used by me at the head of the tables. In 
this place I wish only to mention that the areas as they are defined do not claim to be 
exactly phytogeographical divisions, although I have tried to conform them to tin; 
corresponding divisions as closely as I could witliout unduly multiplying them or 
impairing the clearness of the tables. The great number of endemic species, many 
of which are distinctly representative forms, led me to introduce the sign ^, 
indicating that a species is represented by a more or less closely allied form in another 
area. The same sign was further used in a few cases of non-endemic species, when they 
were represented outside their undoubted area by so closely allied forms that their 
specific autonomy seemed to me doubtful. 

I purposely placed Borneo in the middle ot" the tables and grouped the ot her areas to 
the right and left according to their position and distance ejist or west of Borneo. Thus 
a glance at the tables reveals at once several very remarkable facts, such as the great 
preponderance of the relations with the remainder of Malaya over those with Austro- 
Malaya and the Philippines, the considerable number of species — identical or repre- 
sentative — common to Kinabalu and the Himalaya region, and the comparatively large 
share of what we might call Austral- Antarctic elements. Yet a mere comparison of the 
arithmetical results derived from the tables would l)e greatly misleading in several 
important points concerning the affinities and the history of the flora. In fact any 
conclusions drawn from results obtained by tabulation must be subjected to a careful 
scrutiny before they become valid. Tlie total area of the species and of the natural 
groups of which they are members, the grouping of the members of each group within 
their common area and their phylogenetic relations, the conditions which influence their 
spreading or their extermination, and the geological history of the corresponding part of 
the surface of the earth, are the most prominent points to be considered. I need not 
point out how very little we know at present in regard to almost every one of these 
questions. How^ever, I have tried to ascertain as much as I could from the splendid 
material at Kew and from literatm*e, in a necessarily limited time, and I place the results 
before the reader in full consciousness of their incompleteness. They agree in many 
points with Hooker's researches on the diffusion of Boreal types across the Arcliipelago 
to Australia and of Australian and Antarctic types in the reverse direction, in his classical 
Introductory Essay on the Flora of Tasmania, with the general outlines of the phyto- 
geograpliy of Malaya as laid down in Engler's ' Versuch einer Entwicklungsgeschichte 
der Pflanzenwelt,' and with Warbui'g's latest and important papers on certain parts of 
the Malayan flora, papers which have the invaluable advantage of being based on 
independent research. This coincidence makes me feel more confident than would be 
the case with a different result. 

SECOND SERIES. — BOTANY, VOL. IV. O 



m DE. O. STAPr ON THE I'LOEA OF MOUJNT KINABALU. 

1 think it expedient, for the discussion to which this paragraph is devoted, to divide the 
plants represented in the collections from Kinabalu into three sets : (1) the flora of the 
hill zone (with the exception of the weeds), (2) the flora of the j^rimary forest above 
SOOO feet and of the bush of the summit zone, and (3) the flora of the bogs and rocks. 

A. The Flora of the Hill Zone. 

The three most powerful agencies which cause in the present, and, so to say, before our 
eyes, important changes in the natural associations of plants in the Malay Archipelago, 
reducing or exterminating on one side and giving place to new settlers on the other, are 
the slow hut steady advance of the coast-line wherever it is skirted by a mangrove belt, 
volcanic eruptions, and the clearing of the land from the primary vegetation by man ; the 
last certainly the most extensive of the three causes. It is the only one which comes 
under consideration in this paragraph. The history of the plateau of Dieng and of the 
lo^vland of Balabuan in Java are instances of the comj)arative rapidity with which 
cultivated land may be recovered so completely by the most luxuriant forest growth that 
we should take it for primeval forest never disturbed by man, but for historical proofs of 
their age. The replacement does not always take place, and it is probably in many cases 
more or less incomplete. So far as Kinabalu is concerned, I have no indication that the 
clearings extended at any time much beyond what I have called the hill zone. This 
part of the country, however, seems to be kept in a perpetual state of secondary forest or 
of " young jungle " by the constant shifting of the cultivated area. No wonder that 
not only a number of weeds which generally follow man in this part of the tropics have 
found their way into the country, but also other elements which spread easily and are 
more adapted to the peculiar conditions of cleared land, as the increase of light, the 
decrease of moisture, the reduced stability of the climate, &c. This must ultimately 
change the individual character of the local floras, the more so the longer it acts, 
and replace it by a more general character. Poor as the collection of the hill zone is, it 
nevertheless bears out this deduction. Twenty-one species, or 50 per cent., arc common 
throughout the greater part of Tropical Asia, not a few of them extending at the same 
time to Australia. To these may be added four or five of the endemic species of this 
zone on account of their close affinity to so widely-distributed elements. Eourteen 
other species, or 33 per cent., have a more limited range of distribution ; but this covers 
still the whole of Malaya in eight cases, whilst the species are replaced in West Malaya 
(outside of Borneo) by representative forms in six cases. Three other endemic species 
are more individualized, but still distinctly Malayan elements. The only species 
which remain are BrooJcea albicans and Scyphostegia borneensis. Brookea is a genus 
hitherto found exclusively in Xorth Borneo, having no distinct affinities in the Old 
World floras, and being doubtfully related to a Brazilian genus. The genus Scypho- 
stegia is endemic on Kinabalu and still more isolated, systematically as well as 
geographically. Thus the character of the hill-zone flora is to almost GO per cent, 
generally Indo-Malayan and otherwise Malayan, but slightly imbued with a local 
shading due to the presence of a few endemic species, most of which have no very marked 
characters. 



DE. O. ST.APF ON THE FLORA OF MOUNT KINABALU. ^5 

B. The Plora of the Primary Forest and lUsii above 3000 feet. 
Tliis flora as represented in the collection comprises about 258 Phanerogams, mostly 
trees and shrubs, 32 Perns, and from 6 to SclaglucllaceiV. As it constitutes the p^reat 
bulk of the Phanerogams, we may assume that the proportions of the endemic element in 
the subsequent zones of altitude are about the same as already indicated for the total 
vegetation (see p. 91). Almost exactly two-thirds of the Phanerogams are endemic 
species. The endemism, however, increiises with increasing altitude to a certain point, 
when it becomes stable. It is 47 per cent, from 3000 to 5000 feet (exclusive), G5 per cent. 
from 5000 to GOOO feet (inclusive), and 08 per cent, from above GOOO feet, whilst it was 
not quite 20 per cent, in the hill zone. 

1. Indo- Malay an Elements. 

If we designate all the non-endemic species of the collection, tlie distribution of whieli is 
limited to Tropical Asia, with the exclusion of Arabia and with the addition oi Tropical 
Australia, New Guinea, New Caledonia, and Polynesia, as Indo-JIalayan elements, and 
if we add to them tliose endemic species a\ liicli are representatives of such species, or, 
in any case, distinctly allied to them, we find about 240 species, or 93 per cent., belonging to 
this class, the remainder being Austral- Antarctic (16) or Boreal (2) elements. The Boreal 
elements are herbs from the lowest part of the primary forest, whilst the Austral-Antarctic 
ones are almost peculiar to the highest zones. If we therefore distribute the Indo- 
Malayan elements according to the three zones, we get the following figures : — 

Second zone 97 per cent. 

Third zone 90 „ 

Pourth zone 71 „ 

(i.) Close?' Affinities. 
The Indo-Malayan elements may be classed according to their closer affinities as follows 
below. The assignment of a species to one of these classes is, of course, arbitrary to a 
certain degree, and we cannot expect to find any strict lines to go by. The classification 
should be taken as a whole and judged from its average results, when, I believe, it will 
be found tolerably correct. The fact that a species is found on Kinabalu and in the 
Himalaya does not justify in itself its designation as a Himalayan element or type if this 
name is to have a meaning at all. If the same species or representative species are found 
in Java and Sumatra or in the Malay Peninsula, and if the natural group of which they 
form part has its centre in Malaya, whilst the species stands more or less isolated 
in the Himalayan region, then we may appropriately call the species a Malayan element, 
notwithstanding its extension into the Himalayan region. We are in a similar position 
in many other cases, as will be seen from a glance at the tables, where, for instance, a 
number of species or representative species are indicated for Japan and temperate China, 
althouo-h I have no distinctly Japanese elements to deal with. The truth is that all these 
elements which are indicated for Japan are foreign to the Japanese flora, or so widely 
distributed that they are not specifically Japanese. However circumstantial and subject 
to differences of opinion in detail this way of dealing with the geographical affinities 

o2 



96 DE. O. STAPE ON THE TLOEA OF MOrisT KINABALU. 

of a given flora may be, I think it is the only one which is truly rational, and it never 
can be replaced by the mechanical summing-up of the indications made in the tabu- 
lations. 

a. General Indo-Malaymi Elements (elements distributed throughout Indo-Malaya). — 
I refer 39 species, or 16 per cent., of the Indo-Malayan portion of the flora to this class. 
One-third are endemic, two-thirds more or less widely distributed ; but there are a few 
species among the endemic ones which approach so very closely to well-known species 
that some botanists would probably consider them rather as varieties than as species 
[Pavetta limhata, Ardisia oocarpa, Daphniphyllum horneense). They are pretty equally 
distributed from 3000 to 11,000 feet, but comparatively numerous in the summit zone, 
where they amount to 30 per cent, of the Indo-Malayan portion and are mostly endemic. 
This circumstance, and the fact of their belonging almost exclusively to genera Avhich are 
apparently very ancient {Eurya, Ilex, Myrsine, Daphniphyllum), suggest that they are 
rather remnants of an old flora than that they owe their presence to some particular 
capability of spreading, as this is the case with most of the widely-distributed species of 
the hill zone. 

b. Elements peculiar to the insular portion of Indo-Malaya (including the Malay 
Peninsula). — I find only 9 species which are either known from the insular portion 
of Indo-Malaya only, west and east of Kinabalu, or, if endemic, have representative 
congeners within this limit only, and they would coincide almost entirely with the 
general Indo-Malayan elements if we took the afiinities in a somewuat wider sense. 
There are, in fact, only two genera {I? entapTiragma and Dichotrichium) which are 
absolutely limited to the Indo-Malayan Archipelago. 

c. Malayan Elements. — These amount to 129 species or 50 per cent, of the Indo- 
Malayan portion of the flora and to almost 40 per cent, of the total vegetation, thus 
forming the principal feature in the composiLion of the flora of the forest. They are 
most numerous between 3000 and 5000 feet (60 per cent.), and least numerous in 
the summit zone (45 per cent.). Not quite two-thirds of them are endemic species, 
many, however, closely allied to species from Java, Sumatra, or the Malay Peninsula. 
The increase of endemism with increasing altitude is manifest also here, but it 
decreases again in the summit zone. The percentage is 50 from 3000 to below 5000 feet, 
68 from 5000 to 6000 feet (inclusive), 75 from above 6000 to below 11,000 feet, QQ 
from 11,000 feet. 

d. Bornean Elements. — The species which are limited to Borneo, or, if endemic, have 
no close congeners outside Borneo, number about 34. AU but 6 were collected at or 
above 5000 feet, and all of them are endemic on Kinabalu except four. They belong 
chiefly to Melastomacece, an order in which the flora of Borneo is particularly rich, and 
to the genera Vaccinium, Hhododendron and Nepenthes. Sir Ferdinand von Mueller 
considers, however, some of the species of Rhododendron and Vaccinium described by him 
from the Owen Stanley Bange as very near to species which I refer to this class. 

e. Elements common to Kinabalu and the Philippines only, — I refer to this class 
11 endemic and 1 non-endemic species: 10 of them are found between 5000 and 
11,000 feet, and 1 below and 1 above this zone. The majority of them point most 



DR. 0. STAFF ON THE FLOKA OF MOUNT KINABALU. 97 

distinctly to Malaya, so that we cannot call tlicni Philippine elements at all. The 
remaining 3 species are Evodia mibunifoliolata, Ixhododendrou cuueifolntm, aiul llh. 
ericoides. The two Rhododendrons are represented by one or two very close allies on the 
volcano Ap6 in Mindanao. Sir Ferdinand von Mueller describes Hhododendron 
gracilentum as very similar to Rh. ericoides, but I do not know the species, except from 
description. There may be some resemblance with certain Himalayan species, but the 
affinity is by no means close enough to trace them back to these, or cice versa. They 
form a small group by themselves, characteristic of Kinabalu and Mindanao. Yet it is 
impossible to say whether they are more Philippine or more Kinabalu elements, and, if 
Sir Ferdinand von Mueller's Mliododendron gracilentum is really so closely allied to Bh. 
ericoides, they would be members of a type of far wider and more eastern range. As to 
Ei'odia suhunifoliolata I must refer to the special part. The species is, moreover, too 
imperfectly known to base upon it any deductions. 

f. JElenients common to Kinabalu and Austro-Malat/a only. — I have to mention only 
2 or 3 species as closely allied to Austro-Malayan forms, and having at the same time no 
distinct and close relations otherwise. They are Medinilla lasioclados, Elatostemma Lowiif 
and a species of Myrsine. The specimen representing this Myrsine is too incomplete for 
description, but sufficient to show a distinct affinity with a small Austro-Malayan branch 
of the genus. The Medinilla belongs to a group represented in the cjistcrn part of the 
Indo-Malayan region and even extending into Polynesia. The pliylogenetic relations, 
however, of the species of Medinilla as well as those of Elatostemma are still too obscure 
to justify me in laying any particular stress upon these instances. This almost entire 
absence of 'pronounced Austro-Malayan elements is another very characteristic feature 
in the flora of Kinabalu. Yet it must not be overestimated, as avc shall see that 
other very important relations — relations of a higher order and a more remote date — 
exist between the flora of Kinabalu and Austro-Malaya. 

g. Elements common to Kinabalu and Tndo-China, and belonging to types having a 
centre in Indo-Chhia. — The only species to be mentioned here are an Illicium^ too 
imiDerfectly known to be described, and Stranvcesia integrifolia. The Asiatic area of 
the genus Illicium covers Indo-China and subtropical China, and extends westward to 
the Khasia Hills, Lower Burma, and the Mala^ Peninsula. One species is known 
from Florida, and another from Cuba. It is one of those archaic genera which once 
must have had a Circumpacific distribution of which the present areas are only remnants. 
Whether it had formerly a wider range in the Malay Archipelago, or whether the 
habitat on Kinabalu marks the south-eastern limit to which the genus advanced from 
what is at present Continental Asia, we do not know. The genus Stranvasia is known 
from South-west China, from the Eastern Himalaya and the Khasia Hills. It is 
closely allied to Photmia, from which it differs mainly in a technical character derived 
from the fruit. There are very few species of FJiotinia which reach the Malayan region, 
and all of them can be traced back to the Himalayan region or to Indo-China and even 
temperate China, where they join the true Boreal flora to which the suborder of Fomacece 
almost exclusively belongs. 

h. Elements common to Kinabalu and the JSimalayan Region only. — The number of 



98 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

species which are commoii to, or mutually represented in, both areas is very considerable, 
as can be seen from the tables. But nearly all belong either to the class of general 
Tndo-Malayan or Malayan elements, in the latter case representing outliers of the 
Malayan flora in the Himalayan region. There is one species in the Kinabalu 
collection which is identical with a species from the Himalayan region, namely, Sabia 
parvijiora, and 2 or 3 are representatives of Himalayan species, althougli they have, so 
far as we know at present, no very close congeners in the intervening countries. Yet 
this discontinuation of the areas ceases at once so soon as we take the affinities in a 
"somewhat wider sense. In fact, there is not a single sj)ecies in the flora of the primary 
forest and bush of Kinabalu which could be traced back immediately to the Himalayan 
region as the cradle of the type w ith anything approaching certainty. 

i. Elements common to Kinabalu and the Ceylon Hegion only. — Microtrojyis ramijlora 
has been hitherto known to inhabit Ceylon and the "Western Ghats only. Tlie Kinabalu 
plant enumerated under this name agrees so exactly with certain states or forms of the 
Ceylon specimens that I am not able to distinguish it. In 6 other cases {Eugenia 
Jcinabaluensis^ E. ampullacea, Lasianthus memhranaceus^ L. euneurus, L, rotundatus^ 
and Glochidion teniiistylum) the affinity with species from the Ceylon region is 
strikingly marked in the vegetative character as well as in the floral structure. True, 
there are numerous species of Eugenia and Lasianthus., several of Glochidion 
§ IIemigloch.idio7i ., and a few even of 3£icrot7'oj)is, in Malaya, which are allied in a way to 
the Kinabalu species, but I do not know a single instance of a somewhat closer affinity. 
These 7 species stand pretty isolated amongst their congeners in Malaya. The only 
suggestion concerning them which appears to me rational is to consider these species as 
belonging to very old types, and as relics of a flora which was in a more intimate 
connection with the flora of the Ceylon region. 

(ii.) Broader Affinities. 

It will have been noticed that I used the term '* element " in a very narrow sense in 
dealing with what I called General Indo-Malayan elements, Malayan elements, &c. 
The results derived from the comparison of the primary forest flora of Kinabalu with 
the conterminous floras on this base were very striking in several ways, particularly with 
regard to the almost entire absence of Austro-Malayan elements in the Kinabalu flora. 
I have mentioned, however, already that this result would be very difierent if we based 
the comparison on groups of a higher order — I will call them briefly tyjms— such as 
genera, if they are very homogeneous, or subgenera and sections, if they are not. I 
have tried to ascertain these broader affinities and place the result before the reader 
in the following lines. These types are, of course, spread over wider areas, and 
their geographical classification will therefore be accordingly more general. We may 
divide them into the following classes : — 

a. A^nphiti^op'tcal Types (types common to the Tropics of the Old and New World). — 
These amount to about 14 per cent. They are, with very few exceptions, of little interest, 
as most of them are genera, or sections of genera, consisting of species the natural 
affinities of which are still very obscure. They do not throw, therefore, any particular 



DE. O. STAFF ON TJIE FLORA OF MOUNT KlNAliALU. 99 

light upon the history of the flora. Some are prohahly very old types, as, for instance, 
Myrica ; others may have attained their almost world-wide distrihution at a comparatively 
late date, as Bubiis § Idccohatus; and others, which are already hecomiiiL,^ hettcr 
known, suggest a differentiation more in accordance with their geographical distrihut ion, 
as, for instance, Sterculia § Integrifoluc, which seems to consist of two natural groups, one 
confined to the Old, and the other to tlie New World. 

h. Talceotropical Types (types confined to the Tropics of the Old World generally). — 
These consist of ahout 20 per cent. Not a few* of them have their ])resent centre 
in the Indo-Malayan region, extending westward only to the Mascarenes or to 
Madagascar, or in a few species to Tropical Africa. Thus, Melastotna does not go beyond 
the Seychelles and might better be counted with the Indo-Malayan types. Nepenthes 
extends to Madagascar, but not to Tropical Africa, and has its centre most decidedly 
in the Indo-Malayan region. The same may be said of Adinamb^a, Vygoiirn, Euycnia 
§ St/zyge^wi, Argostemmay Mccsa, Bmhelia, Ard'ma, Alyxia, Strohllanthes, and probably 
of several more. Adinandra, JPygcum, and Argostemma, for instance, are each represented 
only by a single species in Tropical West Africa, thus leaving an enormous gap between 
the two areas inhabited by them. Dicmella is very widely spread, from Polynesia to 
Madagascar ; but it does not reach the African Continent. Most species of tlie genus, 
moreover, being Australian, it would perhaps better be connected with the Austral- 
Antarctic types. If we may venture to make any suggestion as to the origin of these 
types, we must say that all the evidence ascertainable is in favour of the Indo-Malayan 
region. The remaining tyj)es, though represented by more and even by numerous species 
in Tropical Africa also, cannot be said to be more .African than Indo-Malayan. In some 
cases, as Impatiens and Eugenia § Syzygenm, the centre is distinctly more in the western 
part of the Indo-Malayan region, and the relations with Madagascar and Tropical Africa 
are rather strong. On the whole, hoAvever, these types would not justify the assumption 
of an " African " branch in the primary forest flora of Kinabalu. All these types 
extend also east of Kinabalu and to the Philii^pines, with the exception of Argostemma^ 
which does not range east of Kinabalu, although it is known from the Philippines, and 
of Adinandra, which I know from Celebes, but not further eastwards. They are, 
however, more abundantly developed west of the Macassar line than east of it, where 
many of them are represented by a few species only. 

c. Indo-Malayan Types (types confined to the Indo-Malayan region). — I find these 
represented by 68 genera or sections, or by 56 to 57 per cent, in the primary forest flora of 
Kinabalu. 25, or two-fifths, are found only Avest of Kinabalu, 3 extend to Celebes, the rest 
range more or less eastward, many to Tropical Australia or even to Polynesia. Some of 
these number many species in Austro-Malaya, as for instance Medinilla, but the majority 
have their centre most distinctly in the wide region between the Sunda Islands and the 
Himalaya ; and there is not a single genus, subgenus, or section among the Indo-Malayan 
types of Kinabalu which might be considered truly " Austro- Malayan," i. e., having at 
present its centre in Austro-Malaya, and the same is the case with regard to the 
Philippines. Melastomacece, Rubiacece, Ericacece (incl. Vacciniece), Gesneracece, and 
Orchidece are the most prominent orders so far as these types are concerned. 



100 DE. O. STAFF ON THE FLORA OF MOUNT KINABALTJ. 

d. Cii'cumjmcific Types (types principally developed in the Indo-Malayan region, but 
dispersed through the Boreal regions to America). 

Illicium. Rhus § Venenata, 

Ternstroemia. Rhododendron § Eurhododendron. 

Saurauja. Symplocos § Lodhra. 

Ilex § Paltoria, Quercus § Pasania. 

Perrotettia. Castanopsis. 

Smilax § Nemexia. 

Some of these types extend considerably north into the temperate regions, and where 
they range south of the Equator they certainly converge to the north. They are at the 
same time absolutely absent from Tropical Africa and the Austral-Antarctic region. 
Illicium, Rhus § Ve^ienatce, Castanopsis, and Smilax § Nemexia are not known to extend 
east or south-east of Borneo, and only Rhododendron § Eurhododendron and Saurauja 
are represented by a greater number of species in Austro- Malaya. It is very difficult to 
form a more exact idea as to the origin of these types ; but we may assume pretty safely 
that they are very old types, which for the most part developed either in the high 
mountains of Tropical East Asia or in higher latitudes, but still in East Asia, whence 
they spread to America. 

2. Boreal Elements. 

The only distinctly Boreal elements in this forest flora are Viola serpens and Phlomis 

rugosa. Viola serpens has a very wide vertical and horizontal range in the Himalaya, 

ascending to 10,000 feet (Kilar Pangi, North-west Himalaya, Watt), and descending to 

3000 feet, but generally ranging from 5000 to 8000 feet. It descends to 4000 feet in 

Ceylon, and it was collected between 4000 and 5000 feet in Moulmcin. The habitat on 

Kinabalu is therefore exceptionally low. I have seen, however, a specimen from the 

Ulu (upper) Batang Padang Biver, in the Malay Peninsula, stated to have been collected 

at 400 feet. If this indication of altitude is correct — it might be a slip for 4000 feet 

— it would suggest an extraordinary indifference of this species for climatic conditions. 

But, as the Batang Padang Biver drains from a mountain-range rising to 5000 and 6000 

feet in altitude, the specimens collected at 400 feet may have been stray individuals 

carried down accidentally from the subtropical or temperate zone of the mountains. 

This capability of coping with almost any conditions of climate, except over-dryness, 

explains sufficiently the intrusion of this Boreal type into the Malayan flora. Rhlomis 

rugosa differs comparatively more from its Central- Asiatic and Mediterranean congeners, 

and it does not ascend anywhere much into the temperate zone, being confined to the 

tropical or subtropical zones. Its Boreal origin is nevertheless as certain as conclusions 

regarding the origin of a species drawn from the present distribution of the congeners 

can be. 

3. Austral-Antarctic Elements. 

I include in this class 16 species which are more or less closely related to elements 
inhabiting the subtropical and temperate parts of Australia and New Zealand. Most of 



DE. O. STAFF OX THE FLOKA OF MOUNT KINABALU 101 

them extend at the same time to the Antarctic rcjjioii, and throui^li it to temperate 
Soutli America and the Andine region. They may be classitied into 3 groups : — 



a. Angiospermotis Elements extending across the Antarctic lirgion. 

Drimys piperita. Nertera depressa. 

Myrtus flavida. Pratia borneensis. 

Coprosma crassicaulis. Gaultficria borneensis. 
C. Hookeri. 

Drimys piperita seems to be very characteristic of the ridge-vegetation of Borneo and 
of East Malaya, and it is also kno\^Ti from the Pliilippines. The genus is represented 
by several species in New Caledonia, Australia, and New Zealand, and by some — though 
opinions vary much as to tlioir number — in America, where it rangers from tlu* extreme 
soutli along the Andes to Mexico. The nearest allies in Australia grow in humid 
forests up to a considerable elevation, and most — if not all — of the remaining species are 
distinctly hygrophilous. Coprosma crassicaulis and C. Ilookeri are closely allied to 
certain species from New Zealand and South-east Australia, all of llicni being most 
decidedly hygrophilous. This genus as defined at present does not extend to South 
America, though one species is known from Juan Fernandez and another from the 
Antarctic Islands south of New" Zealand. It is re^jlaced in South America by the genus 
Nertera^ which differs from Coprosma only in the mode of growth. Nertera extends 
also over the Antarctic region and Australia and New Zealand to the Malay Archipelago, 
and the commonest species, N. dejjressa, is amongst the Kiuabalu plants. It is also 
typically hygrophilous. Fratia borneensis is a close congener of P. {^Colensoa) physa- 
loides, Hemsl., a herb growing in shady woods in the subtropical parts of New Zealand, 
the two species forming together a very natural section of the genus. This section does 
not extend to America, but another section, Eu-Pratia, numbers several species in 
Extra-tropical South America, whilst the remainder is confined to New Zealand and 
Malaya, except one species peculiar to Malaya and the mainland of Asia as far as the 
Himalaya and South China. A third section, Speirema, is distributed (in a single 
species) from Malaya to the Himalayan region. All these species are, so far as I 
could ascertain, hygrophilous, the small species of § Eu-Fratia growing in wet soil, 
particularly along rivers and streamlets, and § Speirema in shady and humid forests. 
One of the little species of § Eu-Pratia, the Indo-Malayan P. begonicefoliay occurs also 
occasionally as a w eed in cultivation, and I infer the same for P. (§ Speirema) montana 
from a note on a label attached to a specimen of it. Gaultheria borneensis approaches 
closely to G. antipoda, a species inhabiting the mountains of Tasmania and very 
common in New Zealand, where it descends sometimes to the sea-shore. Several more 
species are known from Australia and New Zealand, a few from the Indo-Malayan region, 
and there is the great bulk of the genus spread over America from the extreme south to 
Oregon, but not one of the Indo-Malayan species is by any means so closely allied to. 
G. borneensis as is G. antipoda. Although not dependent on perpetually moist soil or on 
the shady shelter of woods, G. antipoda might yet be considered in a certain sense as a 

SECOND SERIES. — BOTANY, VOL. IV. P 



102 DE. 0. STAPi^ ON THE PLOEA OF MOUNT KINABALU. 

lijgropliilous plant, like so many Gaultheriese dependent on the freqnent saturation of 
the atmosphere with moisture ; and the same is evidently the case with G. boimeensis, 
the anatomical structure of the leaves of which is almost identical with that of 
O. antipoda. 

h. Angiospermous Elements not extending to the Antarctic Region. 



Leucopogon malayanus. 
L. suaveolens. 



Leptospermum recurvum. 
L. javanicum. 

Leptospermum numbers 3 species in Malaya. They are very closely related to each 
other and to L.flavescens, Sm., of Australia (in the sense of the * Plora Australiensis ') ; 
in fact, more closelj^ than the extreme forms referred to X. flavescens by Bentham. 
They range on the mainland as far as Moulmein, and they seem to be confined to the 
drier summits of the mountains. On the other hand, there are at least 20 species in 
Australia and Tasmania, most of them being found in East Australia, besides several 
species in New Caledonia, and 2 in New Zealand. They are closely allied and generally not 
high-level plants, like their Malayan congeners. Some of them grow in moist localities, 
others along rivers ; but, taken as a whole, the genus appears rather adapted to a drier 
and, above all, sunnier climate. Leucopogon is in a similar position. There are also 3 
species in Malaya, whilst not less than 118 species are enumerated in the 'Flora 
Australiensis.' A few occur also in New Caledonia and 2 in New Zealand. If we, 
however, follow Sir Ferdinand von Mueller and sink Leucopogon in Styphelia, the 
number of Australian species would still more increase and the area extend to the 
Sandwich Islands. The fact that more than 80 species are found in West Australia 
suggests still more a particular adaptation to a dry and sunny climate than in the case of 
Leptospermum. This would explain why, for instance, Leucopogon malayanus prefers 
open, airy, and light places on the ridges and summits of mountains, but grows also in 
the low and dry plains of the Island of Phukok. Something similar is the case with 
Leucopogon lancifoliiis. Hook, fil., which seems to thrive equally well in certain sandy plains 
on the coast of North Borneo and on the ridge of the main range of Kinabalu, whilst 
L. javanicus and L. suaveolens inhabit exclusively the highest parts of the mountains. 

c. Gymnosperms. 
I prefer, for several reasons, to treat the Gymnosperms separately. They are 
generally the oldest portion of the phanerogamic vegetation so far as geological evidence 

goes. They play numerically a far more prominent part in the flora of Kinabalu 

and in the flora of the high mountains of Malaya generally — than any other group of 
the Austral- Ant arctic elements, and at the same time a different part. They exhibit a 
very marked habit of growing gregariously, and Podocarpus cupressina appears to 
be the commonest tree of the upper zone of Kinabalu, as it often is in Java at similar 
elevations. They are, physiognomically and biologically, a well- characterized group, and 
their distribution does not quite follow the type either of set a or of set Ij, though they 
are most distinctly members of the Austral Conifer flora. There are 2 species of 
Podocarpus, 2 species of Lacrydium, and 1 species of Phyllocladus. 



J):r. o. staff on the flora of MOUXT KINABALU. lO'S 

Fodocarpus (§ Eii-Podocarjms) hractcata has a very Avide distribution, from liralaya 
northwards to the Himalaya and to Japan. Do Boer indicates it also from Anil)oina. 
Its vertical range is also very wide, extending in the Arcliipelago from the sea-sliorc 
up to elevations of 8000 to 9000 feet, and to 12000 feet in the variety brenfolia. 
It is veri/ closely allied to some other Malayan species, and in a lesser degree to 
P. elata, E. Br., from East Australia, and to P. affinis, Seem., from tlie Piji Islands. 
These species link it more or less to the remainder of the subgenus, whicli is spread from 
New Caledonia to the South Island of New Zealand, to Tasmania and West Australia, 
from temperate South America to the West Indies, and in a few species from the Cape 
to Abyssinia and the Camaroons. Podocarpus (§ Dacry carpus) cnprcssina is one of the 
most characteristic trees of the upper mountain zone of the Malay Archipelago. It 
extends to Mt. Arfak in New Guinea, to the Philippines, and to Upper Burma. There 
are two or three closely allied species in New Caledonia, and one, very common, in New 
Zealand as far as Stewart Island. There is no species of this section known from 
Australia, but a leaf-and-fruit-bearing branch of a fossil specimen, described and figured 
by Baron von Ettingshausen as Podocarpus prcBcupressina from Vegetable Creek in New 
South Wales, seems really to belong to this section. Dacryditim datum is, beside 
D, JBeccarii, which I know but from Parlatore's description, the only species of 
Dacrijdium reported from Malaya. It is represented by a very closely allied, if not 
identical form, in the Eiji Islands, and approaches very much to P. cupressina from 
New Zealand, where the genus numbers 7 species. Four species are endemic in New 
Caledonia, 1 in the Blue Mountains of Australia, 1 in Tasmania, and 1 in South Chili. 
The other species of Bacrydium found oA Kinabalu is still doubtful. Phyllocladus 
hypophylla is a tree or a shrub, extending from Sarawak to Eastern New Guinea, and, 
apparently, common near the ridges and summits of the mountains. Another species is 
found in Tasmania, and a third in New Zealand. The genus seems, however, to have 
extended in the Tertiary period to South-east Australia, as certain fossils described by 
Baron von Ettingshausen evidently belong to it. On the other hand, it should be noted 
that Phyllocladus is the only member of the tribe Taxea^, Eichl,, which is Austral, all the 
other genera being confined to the northern hemisphere. 

C. Flora of the Bogs and Rocks. 

The luxuriant forest -vegetation which, with the exception of the granite cap, covers 
the highland of Kinabalu is interrupted in a few places where water collects, thereby — 
probably in combination with certain conditions of the subsoil — ^favouring the formation 
of bogs, and some rocky precipices whose flora I do not at all know. The principal 
o'ap, however, is caused by the granite cap itself, which excludes almost all vegetation 
except on the top and on the less precipitous southern side above the main range. 
This scanty vegetation of the granite cap is all we know of the rock flora of 
Kinabalu, and to it my remarks refer exclusively so far as they concern the flora of 
the rocks. These bogs and rocks are almost isolated from the surrounding forest- 
veo'ctation by the very conditions to which they owe the presence of a flora of their o\^ti, 

p2 



I04t DE. O. STAPF O^ THE FLOEA OF MOUNT KINABALU. 

and they are at the same time open to a great nnmher of plants from ahiiost any quarter, 
provided that they comply with the special conditions of plant-life afforded by them. 
These conditions, however, are— particularly in the case of bogs — far more universal 
than those of the forest. Yet the absence of competition from the side of the forest- 
vegetation must have another effect, — 'the preservation of remnants of old and retreating 
or already more or less extinct floras, a phenomenon so frequently met with in the bogs 
of the Tropics, as well as of the temperate regions. The position of the rocks is some- 
what modified, as their chemical and physical properties and the modes of water-supply 
are more varied, and the conditions of plant-life, therefore, in every particular case more 
specific. Still they represent the driest, airiest, and sunniest portion of Kinabalu. 
They are exposed, more than any other part of the mountain, to powerful insolation in 
the mornings, and to strong radiation during the nights, thus affording some chance to 
elements of less equable, sunnier, and drier climes. That they are still so poor iu 
variety and quantity may be mainly accounted for by the nature of the granite, which is 
generally unfavourable to luxuriant vegetation, and must be specially so here, as 
the steepness of the slopes does not allow any accumulation of detritus, except in a 
few hollows and crevices. The more universal character of the conditions of plant-life, 
and the absence of any active competition on these bogs and rocks, make it intelligible 
that we find present in these formations plants which often or generally belong to 
different formations in other regions, thus rendering the formations of the bogs and 
rocks much less homogeneous than the forest. In fact, about one-half or even more of 
the bog flora consists of elements which are not typical bog plants, i. e. plants which 
are dependent on the particular physical conditions of bogs, and there is perhaps not a 
single species among those of the rock flora which might be called a typical rock plant. 
There are a few plants in this flora which belong to types which generally participate in 
the formation of the tropical or subtropical forest, and might be considered as elements 
originatuig from this formation and specially adapted to the conditions peculiar to life 
on rocks. But more than 90 per cent, of the flora of the bogs and rocks are perfectly 
foreign to the forest-vegetation. Some of these are Cosmopolitan, some typically Boreal 
or Circumpacific ; but about one-half is closely related to Austral- Antarctic elements. 
AmoDgst these we may distinguish two groups with regard to their broader aflinities — 
one group comprising species which, though closely allied to Austral -Antarctic forms, yet 
belong to Boreal types, and another group which, looked at from the same point of view, 
appears still exclusively Austral- Antarctic. We thus deduce the following classes and 
subclasses. 

1. Cosmopolitan Types. 

Cosmopolitan types are Brosera spathulata, Utricularia orUcuIata, Eriocaulon 
E-Ookerianum, and Scirpus imindatus. They are as species not so universally distributed, 
but are so closely allied to species scattered almost over the whole world that they have 
representatives in almost every one of the great subdivisions of the earth with the 
exception of the Polar region. 




DE. 0. STAPP ON THE FLOEA OF MOUNT KINABALU. 105 

2. Boreal Types. 

a. Boreal Elements. — This subclass comprises those species which are intimately 
connected with Boreal species, namely : — 



Potentilla leuconota. 
P. Mooniana. 
P. parvula. 
Gentiana borneensis. 
G. lycopodioides. 



Car ex vara. 
Agrostis canina. 
Dcttchampsia flexuosa. 
Deyeuxia cpileuca ? 



Pour out of these 9 species are endemic, but one of them, Potentilla parvula., may- 
be perhaps only a state of P. Mooniana (see p. 147). The alTmity of the potentillas 
and gentians is so clear that I need not here discuss it. These types are known 
from the Owen Stanley Range, but they do not extend farther south or east. 
Carex rara, however, ranges as far south as New South Wales, but it is a member of a 
group of hardly separable species which is nearly confined to Japan, China, and North- 
east India. Agrostis and Peschampsia are essentially Boreal genera, and the varieties 
of A. canina and P. jlexuosa found on Kinabalu are, like those from New Zwiland, 
Australia, or South America, only subordinate brandies of Boreal types. Beyeuxia 
epileuca shows perhaps more relations with some Australian species, but 1 am still 
doubtful as to the true systematic position of this plant. Potentilla 2Iooniana, the 2 
species of Gentiana, and Carex rara grow on Kinabalu in more or less boggy places. 
The same Potentilla and the representative species of Gentiana of the Himalaya, 
however, are rather meadow plants than anything else, though the gentians seem to 
prefer or withstand a greater amount of moisture. Potentilla leuconota, Agrostis 
canina, and Peschamima flexuosa are also meadow plants, the grasses generally growing 
gregariously, but on Kinabalu they have become rock plants, nestling in crevices and 
holes of the rocks. 

b. Elements representing Indo- Malay an Branches of Boreal Types. — This subclass 
comprises : 



Aletris foUolosa. 
A. rigida. 
Scirpus Clarkei. 



Carex filicina. 
C. hypsophUa. 
C. fusiformis. 



The two species of Aletris form a small and very natural set which is limited to 
Malaya, but they may be traced back to the more numerous forms which inhabit East 
Asia from the temperate Himalaya to Japan, whilst another species is widely distributed 
through the Eastern States of North America. The genus is a perfectly isolated type ; 
most of the species inhabit boggy or marshy places. Scirpus Clarkei is very closely 
allied to ^S*. subcapitatus, Thw., a species ranging from Ceylon to China, both forming a 
distinct group of their own. Carex filicina is a member of the great group " Indicce,'* 
and particularly of a set which abounds in India and is plentiful in China and Malaya 
Carex Tiypsophila and C. fusiformis belong to a group of closely allied species, 



106 DR. O. STAPE ON THE FLOEA OF MOUNT KINABALU. 

extending from India through China to Japan. All these elements, however, may 
be traced back to types of an essentially Boreal character. Carex filicina and 
Q, hypsophila are known from Malaya proper, but C. fusiformis has hitherto been 
found in the Himalaya only, between 7000 and 12,000 feet. All these species are 
distinctly hygrophilous. 

c. Elements representing Austral- Antarctic Branches of Boreal Types. — These are most 
distinctly related to Austral- Antarctic species, but there is a very remarkable gap if they 
are compared with their Boreal congeners. I refer 3 species to this subclass : 



Ranunculus Lowii. 
Havilandia borneensis. 



Euphrasia borneensis. 



Banunculus Lowii is connected with the Australian group of R. lappaceus through 
JS. amerophyllus from the Owen Stanley Range in New Guinea, but there is no close 
congener among the Banunculi of Asia, Europe, or North America, so far as I am aware. 
Havilandia is an endemic genus, but distinctly allied to that branch of Myosotis which is 
represented by the Austral- Antarctic M. aust^^alis. On the other hand, it approaches 
somewhat to Trigonotis, a genus ranging from North China, Central Asia, and Japan to 
New Guinea. Euphrasia borneensis is in exactly the same position as Banunculus Lowii. 
None of the 3 species is distinctly hygrophilous, nor are their congeners, although some 
of them are at least indifferent to a surplus of water in the soil. 



3. Austral' Antarctic Types. 



Haloragis micrantha. 
Trachymene saniculafolia. 
Drapetes ericoides. 
Patersonia Lowii. 
P. borneensis. 



Cladium borneense. 
C. samoense. 
Schcenus apogon. 
S. melanostachyus. 



Saloragis and Drapetes extend across the Antarctic region to South America. Trachy- 
mene is represented by one species in New Caledonia and by numerous species in 
Australia, but a few allied genera extend from Patagonia along the Andes beyond the 
Equator. Batersonia is also limited to Australia, but it has representative genera in 
Solemanthus and Chamelum in Chile. Cladium borneense belongs to the large group of 
Baumea^ which inhabits Australia, New Zealand, and the islands of the Old World between 
0' and 45° S. lat. from Madagascar to the Sandwich Islands, 3 species only running up to 
China and Bengal. Cladium samoense is a member of the section Vincentia^ which is 
almost exclusively found between 0° and 45° S. lat. in the islands from Madagascar to 
the Sandwich group and Juan Eernandez. The genus Schoenus is, with very few 
exceptions, confined to Australia, New Zealand, and the neighbouring islands, but 
S. apogon is one of these exceptions, being limited to Malaya. ILaloragis seems to 
depend on a certain amount of moisture, though the soil may be stony or sandy. 
Trachymene grows most luxuriantly in wet soil, and in a dwarfed but apparently vigorous 
form in crevices of rocks on Kinabalu. The allied Australian species seem to be equally 
indifferent, and it is in no case a characteristic bog plant. Of the nature of the habitats 
of the Australian species of Batersonia we know very little, but B. glauca, which is 



DE. O. STAFF OX THE FLORA OF MOUNT KINABALU. 10? 

closely allied to the Kinabalu sjiecies, grows in wet and peaty soil in Tasmania, whilst 
others live on sandy ground in generally dry parts of Australia. Schoenus is another of 
those very flexible types which are neither typically hygrophilous nor typically xero- 
philous, but capable of associating with plants of either kind. Drapetes grows, so far as 
I can ascertain, everywhere in open land, evidently depending upon a greater amount of 
light, but at the same time of moisture. 

VI. THE HIGHLAND FLORA OF KINABALU COMPAEEI) WITH THAT OF NEW GUINEA. 
I have already alluded to the analogies in the character of the vegetation of the higher 
parts of the Javan mountains and of the highland of Kinabalu, analogies which find 
their expression not only in the physiognomy, but also in the systematic composition of 
the flora. It would be of very great interest to compare the flora of Kinabalu also with 
the floras of the high mountains of Celebes, Palawan, Mindanao, and Luzon, some of 
which rise to 9000 and even 10,000 feet. Unfortunately we know next to nothing about 
them. On the other hand, Beccari's expedition to Mt. Arfak, Macgregor's exploration 
of the Owen Stanley E,ange, and Zoeller's visit to the Finisterre Mountains have brought 
to light facts which are of great importance for the understanding of tlie flora of 
Kinabalu. 

One cannot read the few lines which Becairi has written on the summit flora ol' 
Mt. Arfak (' Malesia,' i. p. 215) without being reminded of some of the most charac- 
teristic elements of the ridge- vegetation of Kinalmlu. Fodocarptis cupressina, 
P. Rumphii — which Beccari considers identical with P. hracteaia in another place — 
species of JEschynantJms, Diclirotrichmn, Balanophora, Prat'ia, Eunja, Drynus, Lepto- 
spermum, Myrica, and 7 Ericaceae are amongst the plants which his small list of the 
summit flora contains. The physiognomic character of the vegetation of the higher part, 
and particularly of the ridges of the Einisterre Mountains as we know it from Zoeller's 
description and from Warburg's paper " Bergpflanzen aus Kaiser Wilhelm's Land " [in 
Engler, Bot. Jahrb. xvi. p. i. pp. 1-32], seems to be very similar to that on Kinabalu. 
The small collection, however, which Hellwig brought home from this expedition con- 
tains few identical genera if compared with the Kinabalu collections ; yet there are 
among them not less than 5 species of Bhododendron. The highest interest attaches, 
however, to the Owen Stanley Range. Dense forest covers the slopes of this mountain 
up to 11,500 and locally to almost 11,900 feet, followed by a zone characterized by an 
almost purely herbaceous vegetation of " alpine " character, but still of a poor aspect. 
Here, as on Kinabalu, a zone of extreme moisture lies between 7000 and 8000 or 8500 
feet. " Everything is draped in moss, which gives the forest a soft and lonely aspect ; 
and at that time all objects were saturated with moisture." (Macgregor, Journ. Geogr. 
Soc. 1890, p. 201.) " Boots, trunks and branches were wrapped up in thick coverings 
of moss ; even the leaves were not free from it. Everything was soaking wet." (Mac- 
gregor, 1. c. 208.) But this " zone of moss and fog " is followed here by a zone with a 
" dry and fine climate," as is the case in the high mountains of Java. This is very 
different from what we find on Kinabalu, where the granite cap receives an enormous 
amount of rain even during the driest month. 



108 



DE. O. STAFF ON THE FLOE A OF MOUNT KINABALU. 



The following is a list of the plants collected by Macgregor's Expedition, taken from 
Sir Ferdinand von Mueller's paper [Trans. Koyal Soc. Victoria, i. part ii. 1-45], 
compared with the corresponding types from Kinabalu. I may remind the reader that 
Sir P. von Mueller's paper deals exclusively with the plants collected above 8000 feet. 



Macqeegok's Collection. 

Ranunculus amerophyllus. 

Drimys piperita. 

D. hamatensis. 

Hypericum Macgregori. 

Elaocarpus latescens. 

Sagina donatioides. 

Drapetes ericoides. 

Rubus Macgregori. 

R. diclinis. 

Potentilla leuconota. 

Metrosideros Regelii. 

Epilobium pedunculare. 

Helicia Cameronii. 

Galium javanicum. 

Mikania scandens ? 

Anaphalis Mariee. 

Aster (§ Olearia) Kernotii. 

Vittadinia Alince. 

V. macra. 

Myriactis belli diformis. 

Lagenophora Billardieri. 

Ischnea elachoglossa. 

Senecio haplogynus. 

S. erichthioides. 

Taraxacum officinale. 

Vaccinium acutissimum. "^ 

V. Helena. 

V. Macbainii. 

V. amplifolium. 

V. ambyandrum. 

V. parvulifoliuni. 

Gaultiera mundula {Diplycosia) . 

Rhododendron gracilentum. 

R. Loioii. 

R. phceochitum. 

R. spondylophyllum. 

R. culminiculum. 

Styphelia montana [Leucopogon). 



> 



KiNABALTj Collections. 

Ranunculus Lowii. '■^ * 
Drimys piperita. 
D. piperita. -^ 

Elaocarpus sericea. 

Drapetes ericoides. 

Rubus Lowii. -*' 
Potentilla leuconota. 

Helicia erratica. 

[Mikania scandens, a weed of the hill zone.) 



10 species, all o£ the section Epigynum, 



12 species of the genus Diplycosia. 

Rhododendron ericoides. ^ f 

JR. Lowii. 

R. rug sum. '^ f 

R. verticillatum. •^ -j- 

Leucopogon, 3 species. ^ 



* The sign r^ signifies that the Kinabalu species approaches closely to the New Guinea species, 
t According to Sir Ferdinand von Mueller. 



DR. O. STAFF ON THE FLORA OF MOUNT KINABALCJ. 



109 



Decatoca Spenceri. 

Gentiana Ettingshausenii. 

Alyxia semipallescens. 

Veronica Lendenfeldii, 

Euphrasia Brownii. 

Trigonotis Haakei. ^ 

T. inoblita. > 

Myosotis australis. J 

Phyllocladus hypophylla. 

Libocedrus (?) papuana. 

Dendrobium psychrophyllum. 

SisT/rhinchium pulchellum. 

KorthaJsia Zippelii (above 8000 ft. ?), 

Astelia alpina. 

Carpha alpina. 

Scirpus caspitosus. 

Gahnia javanica. 

Schoenus curvulus. 

Carex fssilis. 

Uncinia riparia. 

U. Hookeri. 

Agrostis montana. 

Aira c<2spitosa. 

Danthonia paniculata. 

Festuca ovina. 

F. pusilla. 

Lycopodium clavatum. 
L. Selago. 

L. scariosum. 
L. volubile. 
Gleichenia dicarpa. 

G. flagellaris. 
SchiziBa dichotoma, 
Hymenophyllum tunbridgense. 
Cyathea Macgregori, 
Polypodium punciatmn. 
P. trichopodum. 
Aspidium aculeatum. 
Tanitis blechnoides [Polypodium) , 
Vittaria elongata. 



} 



Gentiana lycopodioides. ~ 
Alyxia lucida and another species. 

Euphrasia borneensis. ■*' 

Havilandia borneensis. 

Phyllocladus hyjtophylln. 

Dendrobium, 2 species. 
Patersonia, 2 species. 

1 species, in the hill zone, figured by Burbidgc. 

Several species. 

2 species. 

Several species, but of other groups. 



Agrostis canina. 
Deschampsia flexuosa. 



Lycopodium clavatum. 

L. Selago. 

L. Wightianum. ^ 

L. volubile. 

1 species. 

1 species. 

2 species. 

Cyathea polypoda and C. Havilandii. 

5 species. 



Polypodium blechnoides. 
Vittaria sulcata. 



There are only six phanerogams in Mueller's list which are also found on Kinabalu ; 
but 22 species, or more than one-third, are represented by rather closely allied forms, 
whilst 9 others are more remotely, but still distinctly related. More than one-half of 
these representative species belong to old Indo-Malayan or Circumpacific types, a few are 
typically Boreal, and about 10 are Austral-Antarctic elements. At least 6 of the latter 
represent, however, Austral- Antarctic branches of Boreal types. Thus it will be seen that 

SECOND SERIES. — BOTANY, VOL. IV. Q 



110 



J)E. O. STAPP ON THE PLOEA OF MOUNT KINABALU. 



the affinity of the floras of Kinabalu and of the Owen Stanley Range rests mainly on 
the presence of certain — and numerous — Ericacese and of a number of Austral- Antarctic 
forms which, however, are mostly Boreal in a wider sense. That the Austral- Antarctic 
element is so much richer on the Owen Stanley Bange is only what we might expect 
from the geographical position of these mountains. The ahsence of composites in the 
highland flora of Kinabalu, whilst they are so numerous on the Owen Stanley Bange, 
is remarkable ; but most of the composites in Sir Perdinand von Mueller's list seem to 
belong to Austral- Antarctic types, only Taraxacum officinale, Myriactis bellidiformis^ and 
perhaps Ana^halis Mari(B being Boreal. 



VII. MEAKS OF DISPEESION AND DISTEIBIJTION. 

I have already pointed to the large proportion of plants producing fleshy fruits in the 
flora of Kinabalu. It amounts to 35-40 per cent., whilst those the fruits or seeds of 
which are apparently adapted by their small ness or certain contrivances for dispersion 
by wind partake with 25 to 28 per cent. I have alluded to the prominent share of 
the endemic element in both classes as a strong reason against the assumption that 
these " means of dispersion " might generally account for the wider or more limited 
distribution of species. I wish to dwell on this point more especially in this place. 

The following table shows the comparative share of those two classes and of the 
remainder, consisting of plants having no particular means of distribution, in the flora of 
the three principal formations. The three columns to the right are added to show the 
proportion of the endemic element in each of these classes and formations. S stands for 
plants possessing fleshy fruits (sarcocarpous plants) ; C for plants having ver^^ minute 
seedsj like dust {coniosjjermous plants) ; P for plants having fruits or seeds provided with 
some flyJng contrivance {jpterocarpous plants) ; and for plants having no particular 
means of dispersion. C and P are represented by a single column, as both are generally 
looked at as special adaptations for dispersion by wind, and as the plants referable to 
P are so very few. 



Formations. 


Percentage calculated from 
total Vegetation. 


Percentage of Endemic 
Species. 


S. 


C,P. 


0. 


S. 


C,P. 


0. 


Secondary forest (hill zone) 


46-5 
45 

50 
59 


18-6 

29 

32 

30 

26 


32-5 

26 

18 

11 

74 


15 
53 
70 
75 


70 
70 
75 
37-5 


30 
46 
70 
66 
24 


Primary forest (2nd zone) 


,, J, (3rd zone) 


„ ,, (4th zone) 


Bogs and rocks 





A glance at this table shows the very different character of each of the three principal 
formations. The secondary forest of the hill zone approaches the primary forest of the 
second zone, so far as the figures in the first three columns are concerned, the difference 



DE. 0. STAFF ON THE FLOKA OF MOUNT KINABALU. Ill 

in class 0, P being mainly due to the increase of coniospermous Mclastoiuacea}, of species 
of Rhododendron and of Orcliidea} towards the upper limit of the second zone. It 
diflPers, however, extremely, so far as the last three columns are concerned. Tliey show 
not only, as I have already pointed out, that the flora of the secondary forest is much 
more universal and that of the primary forest more individualized, even in the lower 
zone (see p. 91), but also that the increase of the endemic element takes place mainly 
among those plants which are apparently best fitted for dispersion. This difference 
still increases if w^e take the third or the fourth zone into consideration. I believe this 
phenomenon is not difficult to explain. 

The land at the foot of the mountain is of comparatively recent formation. It has 
been subjected to the process of clearing to a very great extent and probably for a very 
long time. It shares these peculiarities, and at the same time the general and physical 
conditions of plant-life, with the greater portion of Borneo and with many parts of the 
Archipelago. This process of clearing must have shaken the balance of the aboriginal 
vegetation, weakening it by the alteration of the conditions of plant-life, and at the same 
time opening the way for new competition. The invasion of this new clement was, no 
doubt, favoured by the circumstance that the birds and mammals which might be tlie 
active means of plant-dispersion in this part of the country are commonly species of 
wider distribution, and many of them range freely. Wind must have been also 
more efficient — as a spreading agency — on the cleared land than on the forest-clad 
highland, and finally man himself certainly often became voluntarily or inadvertently 
the cause of dispersion. All this is very different from Avhat we find in the primary 
forest, as we ascend the mountain. Here the land is geologically old land. We have 
no evidence that its vegetation was ever disturbed to any great extent by clearings. 
The physical conditions of plant-life have no longer the same more universal character as 
in the hill zone ; they deviate decidedly and become gradually more peculiar. The vege- 
tative aspect as well as the systematic composition of the primary forest, both the 
result of a long struggle under these very conditions, assume a similarly individualized 
character. The forest forms a harmonious union of congenial elements. Each of them 
is adapted in its way to the nature of the climate and of the soil, and aU are in a state 
of comparative equilibrium which will not cease or materially change so long as the 
conditions out of which it has grown continue to exist. The animal kingdom undergoes 
a similar change as the elevation increases, the endemic species becoming proportionally 
more numerous. Whitehead mentions migratory birds from the coast region of Kortli 
Borneo as well as of Palawan, but not from the highland. True, some birds were found 
there which are exactly identical with species known from the Himalaya, but they are 
not migratory. The fact remains, so far as I can see, that most of the mammals and 
birds of the higher zones — if not all — are confined to the highland, and never leave it. 
Wind may exercise its fall power on the ridges, but not in the high forest which clothes 
the slopes and the bottoms of the valleys, and it can hardly at all affect the ground vege- 
tation, which comprises most of the coniospermous Melastomaceae, Begoniacese, and a part 
of Orchidese. It might carry the extremely light seeds of Bhododendrou or Nepenthes, 
which grow mainly on the ridges, a great distance. In any case, it has not been very 

q2 



112 DR. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

successful in Avidening their areas, as more than two-thirds of the species of these genera, 
found on Kinabalu, are endemic on the mountain. The orchids are in a similar 
position, only two species out of twelve which were collected on the ridge of the main 
range being non-endemic. Thus almost every circumstance seems to be in favour of 
stability and individuality, and contrary to change and exchange. This apparently 
explains sufficiently the extreme endemism of the highland flora of Kinabalu, but it does 
not render the prevalence of sarcocarpous and the great number of coniospermous plants 
more intelligible. It is, a priori, highly improbable that the production of berries or of 
dust-like seeds should not play some prominent part in the life-history of these plants, 
or, as we put it teleologically, not have a distinct design. It is obvious that any con- 
trivance which favours the dispersion of seeds and thereby prevents the accumulation 
of seedlings within reach of the mother-plant is eo ipso an advantage to the species, and 
it is at the same time a well-established fact that these contrivances have in numerous 
cases the desired effect, if I may say so. But this does not necessarily involve dispersion 
over wide areas, and particularly dispersion over areas which are separated by sea or by 
regions of more or less different physical characters. No doubt, dispersion of this kind 
takes place amid favouring circumstances, but it implies the existence of some appropriate 
active means of spreading, be it wind, or birds or some other animals, and its success always 
depends on the presence of conditions favourable to the final settling of the species in 
the new home. Thus the problem is actually very complex, much more so, indeed, than 
often is admitted in speculations on the distribution and migration of plants. I myself 
have no doubt that the presence of many widely distributed elements in the secondary 
vegetation of the hill zone is partly due to the development of particular " means of 
dispersion," as berries, very minute seeds, or flying contrivances, and that it might be 
brought about even under such conditions as now exist. This cannot be said with regard to 
the flora of the primary forest, however favourable the frequent presence of berries or of 
dust-like seeds may have been to wide dispersion under different conditions. 

The third column from the left (in the table on p. 110) contains the figures representing 
the element that does not possess any particular means of dispersion. The figures are 
probably too high, as some of the plants referred to this class may possess some 
contrivance which acts as a passive means of dispersion without our knowing it or being 
able to know it unless by means of observation on the spot. The fruits and seeds of 
others might, although actually not adapted in any way to a particular mode of dispersion, 
be scattered about accidentally. Yet the normal way of widening or shifting their area 
would always be by spreading slowly — so to say, inch by inch or foot by foot. True, 
very vigorous and flexible or plastic types of this kind might — like any others — thus 
push on a considerable distance even in the absence of any powerful impulse from 
without ; but this is certainly quite exceptional, and we may safely say that the present 
distribution of this element is still more unintelligible under the present conditions 
than the distribution of the plants of class S or class C and P. The flora of the 
bogs and rocks, subjected to a similar examination, leads to the same result, although it 
differs considerably from the forest flora so far as the proportional share of the three 
biological classes now under consideration is concerned. There are no sarcocarpous 



13E. 0. STAFF ON THE FLOKA OF MOUNT KINABALU. 113 

plants of any kind among tliem, whilst tlie coniospermous element consists of 37-5 
per cent., leaving thus almost two thirds of the hog and rock plants \n ithout any particular 
*' means of dispersion." Among tlie coniospermous bog plants, Drosera, Utricularia, 
and Eriocaulon are types rather indifferent to temperature. They would— and certainly 
do— succeed in settling in any boggy or generally wet spot between the highest point 
where they w-ere collected and the coast. There is, in the coast region, ccrt-iinly ample 
opportunity for their spreading over wide areas by wind or by ])irds, even under the 
present conditions. Birds are sometimes, as we know from well-established facts, a very 
efficient agency so far as the dispersion of bog, marsh, or swamp plants is concerned, tlie 
seeds easily sticking with mud to their legs or clinging to their plumage. In a similar 
position are most sedges — or perhaps all with the exception of Carex fusiformia and 
C. hypsopMla, which seem to be true high-level plants, and probably also Ilalorag'w 
micrantJia. If we deduct all these plants there remain still about 20 species, b(donging 
either to the rock or bog formation, for the dispersion of which we cannot account by 
merely referring to the general conditions as they at present exist. Four or five of them 
are coniospermous, but among them is only one non-endemic species (Aletrls foliolosa). 
The rest have small seeds or fruits, but certainly not small or light enough to be carried 
a considerable distance by wind. These fruits or seeds may, of course, be occasionally 
scattered by birds or mammals, by w^ater or landslips, &c., but there is, to my know- 
ledge, no agency whatever operating on Kinabalu to bring about in this way an exchangee 
wath those regions where these species or their nearest congeners are found. 



VIII. SUGGESTIONS ON THE HISTORY OF THE FLOKA OF KINABALU. 

I hope to have shown in the preceding section that the composition of the flora of 
Kinabalu points most emphatically to a time when the conditions of plant-dispersion 
were very different from what they are now. The principal circumstances which 
prevent the exchange of the primary highland flora of Kinabalu with other floras 
at present are the absence of birds or mammals, which might be the active means of the 
exchange, and the isolation of the highland. But, as the absence of such animals itself 
is evidently a consequence of this isolation, we are forced to the conclusion that there 
must have been a time w^hen this isolation of Mount Kinabalu did not exist. There 
are two kinds of changes which might put an end to the isolation of the mountain and its 
flora. One is a great climatic change such as would compel the highland flora to descend 
into the hill zone and the plains. It would account for very much ; yet there is no 
evidence that anything of this kind ever has occurred, although a sliffht depression 
of the lines of vegetation of many plants may have taken place. The other chang(^ 
is in the distribution of land and w ater. If Kinabalu once was in immediate connection 
with the highland of New Guinea, or what was then its equivalent, the presence of 
many of the Austral-Antarctic types in the Kinabalu flora would become intelligible. 
A similar connection with continental Eastern Asia would account for the Boreal and 
Ciicumpacific types, and a union of the Malayan islands for the preponderance 
of the Indo-Malayan element. A negative shifting of the shore-line to the extent of 



114 DE. 0. STAPP ON THE PLOEA OP MOUNT KINABALU. 

600 or 700 feet would be sufficient to join Borneo to tlie mainland, to Sumatra and 
Java, and to the Philippines, and it would bring Celebes in reach within a few miles. 
Celebes itself would extend far enough to include Burn, and thus be brought very close 
to Ceram and New Guinea. A shifting of 1600 to 1700 feet, however, would complete 
the union of Borneo with Austro-Malaya, including New Guinea, and with Australia (see 
Berghaus, ' Physik. Atlas,' no. 25). Yet I doubt whether such a change by itself would 
be sufficient to cause a considerable exchange of the highland floras concerned. It would 
raise some of the lower spurs and ranges, with which the highland of Kinabalu is at 
present connected at low levels, to such altitudes that they might exchange a part of 
their present flora with a flora more like the one living now on Kinabalu in similar 
altitudes. Something of this kind would happen in Sumatra or in any other part of the 
united land, and, no doubt, the various highlands would be brought closer together. A 
simultaneous depression of the lines of vegetation such as might have occurred, for 
instance, during a period of extreme glaciation, without seriously impairing the tropical 
vegetation of the lowland, would have favoured the exchange still more. There would, 
however, still have been wide lowlands having a tropical climate and a tropical flora 
where there is now sea, and these lowlands would have been almost as strong a barrier 
for the highland flora as the sea is now. But such a state is hardly conceivable without 
some simultaneous change in the surface of the land. The denudation which takes place 
in the highland before our eyes is enormous. Its configuration and that of the 
surrounding land show clearly enough that it has been going on at this rate for a long 
time. Then the highland must once have been far more extensive, if not also higher. It 
most probably shared this denudation with the remainder of the Bornean highlands, the 
detritus of which is deposited in the vast Tertiary strata which cover one half or more of 
Borneo. Other highlands beside this have very likely existed in the former Malayan 
continent, and, in connection with them, highlands which made the invasion of a northern 
and of a southern element of more or less temperate character possible. In fact, the 
assumption of such highlands seems to me the conditio sine qud non for under- 
standing the history of the flora of Kinabalu and of the highland flora of Malaya 
generally. 

A more exact determination of the lines over which, and of the time when, the 
exchange took place is hardly possible at present. It could be undertaken only on the 
basis of a comparative study of the entire highland flora of the Malay Archipelago and 
of a more complete knowledge of the geology of the region. I restrict myself, therefore, 
to a few general remarks referring to this question. 

Beccari (' Malesia,' i. pp. 219, 220) was of opinion that the Boreal alpine element on 
the high volcanoes of Java must have immigrated at a very recent date, as the volcanoes 
themselves are so very recent. Their habitats on these volcanoes are, of course, as recent 
as or more recent than the volcanoes. But this does not necessarily imply that they did 
not already exist in this part of Malaya before the volcanoes were thrown up. We know 
that the Javan volcanoes rest upon Tertiary strata, and traces of the old Schistose Forma- 
tion which occupies so much space in Sumatra were discovered only some ten years ago. 
The recent volcanoes of Sumatra, on the other hand, rest upon this Schistose Pormation^ 



DK. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 1]5 

and under conditions Tiliicli prove that the volcanoes broke througli a completed, or nearlv 
completed, chain of folded mountains. This old highland may probably have contained a 
great number of such types, which would have been taken over by the volcfinoes as tliey 
were thrown up and the highland broke down. That a Boreal element existed in Malaya 
before the appearance of these volcanoes is clearly demonstrated also by the presence of 
so many Boreal types in Australia. There is no doubt that most, if not all of them, reached 
Australia by way of Malaya and Austro-Malaya, and they can have reached Australia, 
in my opinion, only by w^ay of a land-connection, for— quite apart from other reasons— 
the majority of these types have no particular " means of dispersion." If so, these 
types must have existed in Malaya before the eastern portion of the Indo-Malayan rcion 
was separated from the western. We do not know exactly when this separation took 
place or w^as completed, but it seems to have existed at least since the earlier part of the 
Tertiary period, or long before the volcanos of Sumatra and Java were built up. This 
separation fixes also, on the other hand, the time of the invasion of tlie Austral-.Vntarctic 
element. It must have reached Borneo previous to the separation, although it may 
have spread later northward and westward The next temperate (or subtropical) 
element consists of the Circumpacific types. Their general distribution suggests that 
they are very old types. In which part of Eastern Asia tliey originated I do not venture 
to speculate. In any case, they, or at least some of them, must have been already in 
Malaya before it was separated from Austro-Malaya, as their extension into Austro- 
Malaya proves. Beccari (' Malesia,' i. p. 224) points also to the fact that the line over 
which the Austral- Antarctic types are distributed in Indo-Malaya lies over New Guinea, 
the Moluccas, Borneo and Sumatra, while Java has scarcely any. I lind tliat Java is 
just as rich in these types, or rather just as poor, as Sumatra ; but North Borneo is 
certainly very much richer than both of them. It might be suggested tliat the climatic 
conditions of North Borneo are more like those of the Austral-Antarctic regions. 
This, however, could hardly apply to the "mist zone" of the Sumatran and Javan 
mountains, though it might to the summit zone, which seems to be much drier in Java 
than it is on Kinabalu. Yet, if w^e remember that the Philippines also possess a very 
remarkable set of Austral- Antarctic types (see Bolfe, in Journ. Linn. Soc. (Bot.) xxi. 
p. 296), it seems much more plausible to assume that the main line along which this 
Austral- Antarctic element moved lay rather over Halmaheira and North Celebes than 
over or south of the Banda Sea. 

Every fact that applies to these three components of the temperate highland flora 
applies with still greater force to its fourth and last component, the Conifera?, and to the 
greater part, if not to the whole, of the intra-tropical flora. Both elements must have 
ranged over both portions of Indo-Malaya, east and west of the Moluccas, before the 
seprration of Malaya and Austro-Malaya began, and they did so probably a very Ion- 
time before. 

An interesting question arises with regard to those Austral- Ant arctic types which I 
have designated as a branch of Boreal types of higher order, and I might perhaps add 
to them the varieties of Agrostis and JJeschampsia, and probably also Deyeuxia. A 
remarkable fact is that thev are closely connected with southern forms, but slightly witli 



116 DK. 0. STAPP OA' THE PLOEA OP MOUNT KINABALU. 

their nortliern congeners, although everything seems to point to a Eoreal origin. 
Are they survivals from the great invasion to which the presence of so many Boreal 
types in Australia and even in the Antarctic region is due, or is their presence on Kina- 
balu and on the Owen Stanley Hange the result of a partial re-migration of Boreal 
types from the Austral- Antarctic regions ? In the latter case the phylogenetic gap in 
their relations to their northern congeners might be explained by the transformation 
which the southern branch underwent after having reached the Austral- Antarctic regions. 
We might, however, just as w^ell admit the other alternative under the perfectly justified 
assumption that the northern congeners deviated more from their ancestors than the 
southern ones, which w ere hardly transformed at all. It has also been suggested (Wallace, 
' Island Life,' p. 490) that types of this kind might form the most advanced posts of a 
migratory movement of an extraordinary extent, namely from North America by way of 
the Andes to the extreme south of South America, and thence to the Antarctic region, to 
Australia and New Zealand, and finally to Malaya. Yet the particular types to which 
1 refer in this place do not show a closer afiinity to North American forms. They stand 
in the same abrupt relation to them as to their congeners in temperate Eurasia. 

I will now try to summarize briefly the conclusions to which I come with regard 
to the history of the flora of Kinabalu. The highland of Kinabalu formed, at 
a period not later than the earliest part of the Tertiary epoch, a portion of a much more 
extensive highland, which was — simultaneously or in consecutive periods — connected 
with other highlands which stretched over a wide continental area extending from the 
south-eastern part of the present mainland of Asia into the Austral- Antarctic regions. 
The flora of the then highland of Kinabalu contained already all the essential elements 
which at present characterize it. At the end of this period the highland connection with 
the south-eastern portion of this continent was severed, so that no important exchange 
of temperate elements could take place any longer in this direction. Lowland connection 
may have continued still for a time, after w^hich it also ceased. Up to this point the 
flora of Kinabalu and of the neighbouring parts of the old continent, east and west of 
Kinabalu, formed a natural unit, grown up under the influences of a common history, 
and still manifest in the broader aflinities of the present floras of these areas. After the 
separation of Austro-Malaya a new^ period in the evolution of the floras east and west of 
the line of separation began. The differences which were, no doubt, already indicated 
now became manifest and more accentuated. The highland of Kinabalu was still in 
close connection with the remainder of Malaya and with the Philippines. Then these 
broke away, whilst the connection with West Malaya still continued, and under 
conditions which caused a considerable general uniformity of the flora of Malaya. The 
pronounced Malayan character of the flora of Kinabalu dates from this period. Then 
the Malayan land subsided or broke down, leaving an archipelago. These conditions 
went on and reached a maximum which must have been considerably above the present 
state. Borneo itself was reduced to a group of islands, of which the highland of Kina- 
balu w^as one. The increasing isolation must have acted as a most powerful agent in 
the evolution of the flora. It could not add any new element to the flora, as it tlien was, 
but it gave it that strongly localized or individualized character which still distinguishes 



DE. 0. STAFF OX THE FLOHA OF MOUNT KINABALU 117 

it. Then the shore-line shifted again the other way. The Bornean islands were joined 
into one. The floras of these islands, advancinir witli the advaneiiiij: coast, met, and a 
fresh exchange took place, which, however, coukl not materially aifect tlic highland 
floras, for they were still separated hy the lowland. Tlic exchange and the ensuing 
generalization of the flora remained therefore confined to the lower zones, whilst the 
highlands maintained their individual character. In still more recent time man became 
an agent in dispersion and in generalizing the character of the flora. Large parts of the 
" old jungle " fell by knife and fire, and although the original vegetation gen<M%-illy springs 
up again and again with apparently unabated vigour, there can be no doubt that many 
of its feebler elements have already succumbed in this struggle. Others better adapted 
to the new conditions took their place, and, often favoured by particular means and con- 
ditions of dispersion, spread far and wide. A " secondary " vegetation thus replaced the 
primary or aborginal one, a change in which the hill-zone flora of Kinabalu shared 
equally with the flora of any part of the lowlands or the hills where man settled. 

The last point with which I wish to deal concerns tlie bearing of tli(» results of the 

examination of the flora of Kinabalu on the general differentiation of the Indo-^Ialayan 

flora. The flora of Kinabalu is a new proof of the essential unity of the Indo-Malayaa 

flora west and east of Kinabalu or the Macassar line. But the flora of the Austro- 

Malayan division differs at least as much from that of the Malayan division as the latter 

diff'ers from the flora of the Himalaya region or of Indo-China. True, the relationship of 

the Malayan and Austro-Malayan floras is very much closer if we take into considtjration 

the floras of the coasts, lowlands, and hills, and they become almost uniform if we confine 

the comparison to the shore vegetation. Yet this is the least stable portion of the flora, 

and therefore teaches us almost as little about the history and the affinities of the flora 

as a comparison of weeds would do. We have therefore to rely on the most conservative 

portion of the flora, the flora of the primary forest, and particularly of the highland 

forest. I drew the line between Malaya and Austro-Malava in accordance with Wallace's 

Macassar line, chiefly because it almost coincides with the meridian of Kinabalu. If I 

had included Celebes in ^lalaya, very few species would have been added to the list of 

the Malayan elements of the primary flora. It must, however, not be forgotten that very 

little is yet known of the primary flora of the higher parts of Celebes. This remark applies 

more or less to Burn, Halmaheira, Ceram, and the other islands west of the line which 

separates Warburg's "Papuasian" flora from the flora of "^lalesia" (Warburg, in Engler, 

Bot. Jahrb. xiii. (1891) 230 et seqq.). It is very probable, for geological reasons, that the 

line separating the Malayan and the Austro-Malayan floras will be found to lie much 

nearer to Warburg's line than to Wallace's ; but I believe its exact course will appear to be 

almost as indefinite and practically arbitrary as, for instance, the line which separates the 

Malayan and the tropical Himalayan floras. There will always be a wider or narrower 

strip of borderland between the two divisions. Where it consists of recent land, and 

particularly of land of low ele\-ation and of cleared land, it will bear the more universal 

character of the insular Indo-Malayan flora ; where it comprises primitive and undisturbed 

land it will naturally be rich in endemic elements, and it will approximate more or less to 

the tvpical Malayan or to the typical Austro-Malayan character, according to the position 

SECOND SERIES. — BOTANY, VOL. IV. E 



118 • DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

east or west of the line of geological separation, provided that this line is a definite 
line. This is evidently not the case. The line along which the old Indo-Malayan 
continent first broke up into an eastern and a western portion was certainly not a geome- 
trical line, nor was it formed simultaneously in all its parts. The breaking-up of the 
continent most probably took place over a broad strip of land along different and 
changing lines, and it was certainly not completed before the lapse of a very long time, 
nor has the region in question since been in a state of undisturbed stability. 

I have purposely stated the facts and drawn the conclusions exclusively from the 
standpoint of the botanist, referring only secondarily to the geology of the area. The 
result is essentially the same to which Warburg came, mainly from the examination 
of the "Papuasian " flora. He has already pointed out, and I agree perfectly with him, 
that there is nothing in this theory of the evolution of the Indo-Malayan insular flora 
which is in conflict with the zoo-geographical facts as they are known from Wallace's 
researches and from the more recent publications on the subject, nor do I find anything 
in it which contradicts the results of the geological explorations in the area concerned 
so far as I am acquainted with them. 



IX. TABLES SHO>yiXG THE DIST11TBUTI05 OF THE SPECIES OF EACH OF THE FOUR ZONES. 

(See also p. 93.) 

The geographical divisions adopted in the Tables are : — (1) Boreal Regions : the tem- 
perate regions of the northern hemisphere, with the exception of temperate China and 
Japan. (2) China, Japan : the temperate parts of China and Japan. (3) Indo • China : 
Siam, Erench Indo-China, and tropical and subtropical China. (4) Himalaya : Tropical 
Himalaya, Assam, Khasia Hills, and south as far as Tenasserim. (5) Ceylon : Ceylon 
and the Western Ghats. (G) West Malaya : Malay Peninsula, and the Malay 
Archipelago to the Macassar line, excluding Borneo. (7) Borneo. (8) The Bhilipjrines. 

(9) Austro-Malaya: the Indo-Malayan Archipelago, east of the Macassar line. 

(10) Polynesia (including Xew Caledonia). (11) Australia (including Tasmania). 
(12) New Zealand. (13) South America. 

The sign + signifies that the same species is common to Kinabalu and the division 
for which the sign is entered, whilst the sign ^ indicates that the species is represented 
by a close congener in the division concerned. 

The names of the species described as new in the present paper are denoted by an 
asterisk. 



DR. O. STAPr OX THE FLOKA Oi' MOUNT KINABALU 



119 



A. Hill Zoxe (up to 3000 feet). 



• 




li 






« 




Altitudes on 


8 

B 




'S 


A 




i 

•r 


"a 


'^ i^ ^ 


'a 


c 


je 


c 


Kiiiubiilu. 


a. 


OJS 


B 


9 


J 




a; 

o 




W 1 


t' ! 








=M 






1 




'r 




j 










feet. 










Polygala venenosa 




-^ 


-^ 


-^. 


4- 


4- 


2000-3400 


+ 








j 


*Garcinia Havdlandii 










-^ 


. — . 


2000 












, 


Adinandra excelsa 


> . . 


, ^ 




, , 


— . 


4- 


2000 














*Saiirauja amoena 










— . 


-«-. 


2500 


-^ 


-^ 










Triumfetta suffruticosa 


-^ 


.^. 


. — . 


-^ 


4- 


4- 


2500 


. . 


4- 










*Evodia tenuistyla 




-^ 




-^ 




-^ 


2000 


-^ 












*Salacia laurifolia 






-^ 


— 


-^ 


-^ 


2000 


^- 












Cissus hastata 


1 




, , 


4- 


+ 


2000 












Guioa pleuropteris ': . 










+ 


+ 


20U0 












*Bauhmia excurrens . 




-v^ ■ 






.-^ 


-^ 


2000 


-^- 


"^ 










* Burbidgoi . 








, . 


•^ 


4- 


2000 














Pithecolobium bigeminura . . 1 . 




+ 




+ 


4- 


4- 


2000 


4- 












Eiibus rosaef olius [ 




+ 


4 


+ 


4- 


+ 


25OU-;i00O 


4- 


4- 




4 






Dichroa febrifuga 




+ 


+ 




4- 


+ 


2000 


4- 


+ 










Itea macroph}'lla 






+ 




4- 


+ 


2000 














Decaspermum panlculatum . . 




■^ 


— 




4- 


+ 


2000 


4- 


+ 


■^ 


^- 






Sonerila maculata 






4- 






+ 


2500 














Dissochasta annulata 










+ 


4- 


1200 




4- 










hirsuta 












+ 


2000-3500 














*Kibessia tessellata i 










-^ 


•^- 


2000 














Memecvlon loevigatum i 




■^ 


-^ 




4- 


4- 


1000-2000 


^ 












Mussaenda frondosa, 






+ 


+ 


4- 


+ 


2500 


4- 


4- 










* coccinea 












4- 


2000-3400 














Handia donsiflora I 




4 


4- 


4- 


4- 


+ 
4- 


2000 
1000-2000 


4- 


•f 




4- 






Psychotria malayana 


Campanumoea celebica 




+ 


+ 




4- 


4- 


2500 




4- 










Cletbra canescens 




-^ 






4- 


4- 


2000 


-^ 


4- 










Leuconotis anceps 

Willugbbeia firma 


• 








4- 


+ 
4- 


2000 
2000-3000 














Parameria gland alifera 




+ 






4- 


4- 


2500 


4- 












Hoya coronaria 










+ 


+ 


2000-3000 














Buddleia asiatica 


. • 


+ 


4- 


4- 


+ 


+ 


2500 


4- 


+ 










*Brookea albicans 


. 










''^ 


2500 














^Eschynanthus tricolor 

Thunbergia fragrans 










-h 


4- 


2000-3000 














. . 


-f 


+ 


4- 


4- 


4- 


2500 
2000 
2800 


4- 


4- 


+ 


•• 


4- 




*Strobiianthes kinabalueiisis . . 










■ — • 
















Geunsia farinosa 










+ 


-t- 


+ 


+ 










Polygonum chinense 

Nepenthes phyllampbora .... 




4- 

+ 


4- 


4- 


4- 
+ 


+ 


2500 
2500 
IbOO 
2500 
2000 
2500 
J 600 


4- 

4- 


+ 


-t- 








Seyphostegia borneeusis 

*Lorantbus estipitatus 

Antidesma Moritzii 

Acalypha stipulacea 


.. 


+ 


~ 


~ 


4- 
4- 
+ 


+ 
4- 
4- 
+ 


+ 
+ 
4- 


4- 










Schizostachyum, sp 

Alsophila latebrosa 




+ 


4- 


4- 


+ 


2000 
2000 


4- 


4- 










Davallia ciliata 

tenuifolia 

Asplenium squamulatum .... 




+ 


4- 




4- 

4- 


+ 

- + 

+ 
+ 


2000 

2000 

2000-3200 


+ 
+ 


4- 


+ 






. 


borneense 

Angiopteris evecta j 

Equisetum debile 


.. + 


+ 
+ 


+ 
+ 


+ 
+ 


i: 


2000-3O00 
2000-2500 


4- 
4- 


4- 


: 


4- 







r2 



120 



DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 
B. Lower Mountain Zone (from 3000 to 6000 feet). 






Illiciiim, sp 

Michelia (?), sp 

(jroniothalamus dispermus 
•^ stenopetalus 



Mitrephora 3Iain^ayi .... 
*Melodorum kinabaluense 

Yiola serpens 

Polygala venenosa 

Saurauja ferox 

*Sterculia translucens . . . . 

Impatieus platyi)hylla . . 
*Evodia subunifoliolata . . 

Dysoxylon caulitlorum . . 
*Gomphandra lysipetala . . 

prasiaa 

Perrotettia alpestris . . . . 

Sabia parvidora 

lliibus glomeratus 

rostet'oliiis 

I'raxinifolius 

Drosera spathulata 

*l)ecaspermum Vitis-Idsea 

*Myrtus flavida 

*Tristania elliptica 

Melastoma decemfidum . , 

Beccarianum 

Blastus Cogniauxii 

*Anerincleistus cordatus . 

*Driessenia glanduligera . 

* microthrix 

Sonerila tenuifolia 

* kinabaluensis . . . 

* pulchella 

borneeusis 

*Phyllagathi8 elliptica . . . 

* imitiora 

Marumia pacbygyna . . . 
Dissochajta birsuta 

*Medinilla urophylla . . . 

* lasioclada 



+ 



+ 



+ 



+ 



+ 



-f 



+ 



+ 



+ 



*Anplectruni bomoeandnnu . . . 

Memeoylon lanceolatum 

*riethiaiidra Hookeri 

Casearia leucolepis 

*Begonia adenostegia 

* oblongitblia 

borneensis 

* inostcgia 

Hydrocotyle javaniea 

Artbropbylluui diversifolium 
Weudlandia paniculata .... 

*Argostemma gracile 

* braobyantherum .... 

*Hedyotis protrusa 

*Mussoenda coccinea 

I Acrantbera Atropella 



+ 



+ i + 



+ 
+ 
+ 

+ 



+ 



+ 
+ 



+ 



+ 



+ 



4- 



+ 



+ + 



+ 



+ 



+ 

+ 



+ 
+ 



+ 



+ 
+ 
+ 
+ 



+ 
+ 



+ 



+ 

4- 
+ 



+ 



Altitudes on 
Kinubalu. 



feet. 

40U0 

5500 

3000 

3000 

6000 

3000 

3000 

3000 
2000-3400 

3000 

5000 

5000 

6000 

3200 

6000 

4500 

3u00 

5000 

6000 
2500-3000 

3000 
5000-5500 

5500 
5500-7700 

5500 
5000-5500 

3500 
4000-6000 

3500 

5000 

3500 

3500 

3200 

6200 

3000 
4000-5000 

6000 

3000 
2000-3500 

3500 

6000 

6000 

6000 

4500 



a 


e3 




. f^ 






o^ 


V 


a 


Si a 




h 


|S 




3 


<1 


Ph 


-fj 






+ 



4500 

3000 

6000 

6000 

3000 

3500 

3000 

3500 

3000 

5500 
2000-3400 
5000-6000 



+ 



+ 



+ 


4 


-f 


4- 


+ 


+ 


+ 


+ 



+ 



+ 



4- 
4- 



+ 



4- 



+ 



DE. O. STAFF O^ THE FLOKA OF MOUXT KJNA13AI.L 
B. Lower Mountain Zone (continued). 



121 



i 






e! 
O 

i 
rs 

c 


i 

X 

E 


c 

o 

1- 


« 


2 


Altitude!" on 
Kiuubiklu. 


i 

a 


el 
•< 


a 


« 

1 


1 

1^ 


r 


















feet. 














*Mvrioneuron borneense 
















3U00 




,^ 










*Urophyllum lineatum 
















5500 












*- salicifolium 




.. 










-^^ 


GOOO 












*Ixora kiiiabaluensis 












— 




50(K) 












*Pavetta limbata 






^ 


^ 


-- 


^ 


^ 


6UO0 
3200 


^ 


~ 




~ 






*Psychotria gvrulosa 




* iteophylla 








-^ 








3500 














*Streblosa urticina 








^ 




->„ 




3200 














*Go3rtnera rufinervis 








, , 




-^ 




(JOOO 














*Ohasalia gracilis 












-^ 




3500 




! 










*Lasianthus kinabaluensis . . . 








-^ 




"^ 




3000-32U(» 




1 










* raerabranaceus 










— , 


, , 




3000-5000 




' 










Pratia borneensis 




. , 










+ 


3000 










-.^ 




*Pentaphragma aurantiacum . . 






. . 


, , 




-^ 


-^ 


(j<i00 




^. 










Vaccinium, sp 






, , 


, , 




-^ 


-^ 


3000 














* pachydermum 






, , 


, , 




-^ 


, , 


5500-10,500 














* micranthorum 






, , 


, , 




^. 




a.ioo 














*Diplycosia pinifolia 


, , 








, . 






:\')in) 














* penduliflora 


, , 








, . 




^ 


5(»00 














* urceolata 








, , 




, , 




0000 














* meraecyloides 






, , 








-^ 


6000 














lihododcndron crassifolium . . 












-^ 


-^ 


5000 


^ 












* stenophyllum 














— . 


5000-bOOO 














* durionitolium 












-^ 


+ 


4000-7000 


•^ 












* lacteum 












+ 


+ 


3500 
5500 








-^ 


'■^, 




Leucopogon malayanua 




Majsa indica 






+ 


+ 


+ 


+ 


+ 


4000-5000 
5500 


+ 


+ 










Mvrsine, sp 




Labisia pothoina 












+ 


+ 


3000-4000 














*Embelia phaeadenia 










•-^- 


-^ 




6000 


— 












*Ardisia oocarpa 




. , 


-^ 


-^ 








3000 














* bracbythyrsa 












-^ 


-" 


3000 














• virens 







+ 


+ 




•^ 


•"^ 


3500 
6000 














* amabilis 




oxypbylla 








, , 


. . 


+ 


+ 


5000 














villosa 


, , 




, . 


, . 


, , 


+ 


+ 


4500 














Diospyros, sp 














— 


3000 




! 








, sp 












. . 


-V, 


3000 


! 


1 








*Svmplocos adenophylla 












+ 


+ 


5500-7000 




1 








* Iffitcviridis 


. , 






-^ 








4000 












* Johniaua 


, , 










■^ 




6000 












Rauwoltia serpentina 


. , 






+ 


+ 


+ 


+ 


3000 












Alyxia, sp 


, , 


. . 


, . 






— . 




5000 












Solanum biflorum 






+ 


+ 


+ 


+ ' 


+ 


5000 












ITtricularia orbiculata 


• • . 






+ 


+ 


+ : 


-r 


r.r,(»o 




1 








*-^schynanthu8 magnifica .... 












-^ 


-^- 


.11) 














*Dichrotrichum bracteatum . . 
















.jy^M) 














*Chirita areolata 
















5000-7000 














Cyrtandra Burbidgei 

trisepala 










-^ 


-^ 


■ — ■ 


5000 














. , 












+ 


3000 














*Strobilanthes galeopsis 


. . 














5000-7000 




-^ , 








*Premna cauliflora 
















3000-3200 




( 








*Clerodendron kinabaluense . . 


. . 










-^ 


■^^ 


3200 












Pblomis rugosa 






, . 


+ 




+ 


+ 


3500 


+ 










jN^epenthes Edwardsiana 

Rajah 










. . 






4000-9000 
























+ 


4000-7000 














Peperomia Candida 






• • 


• " 


• ' 


+ 


+ 


6000 
















122 



DR. O. STAFF ON THE FLORA OF MOUNT KINABALU, 
B. Lower Mountain Zone {continued). 



w 



o 



Chlorantluis braehystachyus 

*Litsoea cauliflora 

*Lindera grandifolia 

Helicia erratica 

*Glochidion tenuistylum . . . . 

Baccavirea lanceolata 

Antidesina auritum 

*Claoxylon paucitlorum . . . . 

Aealypha caturus 

Laportea stimulans 

sp. 



+ 



+ I + 



+ 



*Pilea pterocaulis . . . . 

* crassifolia . . . . 

*Elatostemma lineare 

* Lowii . . 

* thalictroides . . 



Burmannia longifolia 
papillosa 



Microstylis calophylla 

*Platyclinis corrugata 

*Bulbophyllum montigenum 
* breviflorum 



*Dendrochilun[i conopseum 
Eria Kiiigii 

* angustifolia 

ferox 



+ 



.S ^ 



Spathoglottis aurea 
*Calanthe ovalit'olia . . . 
*Broinbeadia rigida . . . 
*Appendic'ula congesta 

Alpinia, sp 

*Patersoiiia Lowii . . . 
*Aletris foliolosa 

Smilax odoratissima , 
leucophylla 



Dianella eiisi folia 

*Eriocaulon Hookerianum . . 

*Cladium samocnse 

* borneense 

Scha;nus melanostachyus . 

Miscanthus sinensis 

Dinochloa, sp 

Gleichenia circinata 

Hymenophyllnm dilatatura 

Trichomanes rigidum 

apiifolium 

1 ^ 

' tnaximum 

Pluma 

trichophyllum 



+ 
+ 

+ 
+ 



+ 



Lindsaya jamesonioides 

flabellulata .... 

Asplenium borneense . . 

dichotomuTD . . . . 

" biserinle 

latifolinm 



+ 



+ 



+ 



+ 



+ 



+ 
+ 



+ 



+ 



+ 



+ 

+ 



+ 



+ 
+ 

+ 



+ 



+ 



+ 



+ 



Didymochkiena luuulata 
Gymnogramrae avenia 



+ 



+ 



+ 
+ 
+ 

+ 
+ 



+ 



+ 



+ 
+ 



Altitudes on 
Kinabalu. 



+ 



+ 1 + 
+ 



+ 
+ 



+ 



+ 
+ 

+ 

+ 
+ 



+ 
+ 

+ 



+ 
+ 
+ 



+ 
+ 
+ 



feet. 
6U00 
5000 
3000 
5000 
4000 
3000 
4000 
3000 
3000 
5000 
3000 
3000 
3000 
5000 
5000 
5000 
5000 
3200 
3000 
5500 
6000 
6000 
5500 
6000 
6000 
6000 
4000 
3000 
6000 
4000 
5000 
5000-5500 
5500-7700 
3000 
5500 
5500-8000 
5500-7700 
5500 
5000 
5500-11,000 
5000 
3000 
5000-10.500 
5000-7000 
5000-7000 

6000 
5000-6000 
5000-6000 
5000-6000 
5500-9000 



< I P-i 






■B S 



+ 



4- 



-- I 



+ 
+ 



+ 
+ 



2000-3200 

3000-5000 

3000 



3500 



+ 

+ 



+ 
+ 

+ 
+ 



+ 
+ 



+ 



+ 



+ 
+ 
+ 

+ 

+ 



+ 



+ 



+ 



+ 



+ 



+ 



+ 
+ 
+ 



+ 



+ 



+ 



DE. 0. STAPE ON THE ELORA OF MOUNT KINABALU. 
B, Lower Mountain Zone {continued). 



123 





O 

1"^ 


c 

3 5- 


OS 

.E cs 
O -2 


o 


1 


^ Altitudes on 
2 Kinabalu. 


s 


X _2 


a 


1 


B 


J' 

o 




aa 


6" 


rS S 


'8 


1^ 1 


^ 
^ 


IS 








1^ 












feet. 










i 


Acrostichum bicuspe 






-h •• 




+ . 


4000-8000 




+ 










Schizfea malaccana 












+ . 


5000 


+ 












Lycopodium serratum 






+ 


+ + 




+ + 4000 


+ 


+ 


+ 








Selaginella caulescens . . 






+ 


+ + 




+ + 3000 


+ 


+ 


















+ + 




+ + 3000 


. , 


+ 
















Dicrannm assimile 










.. + 




+ . 


6000 














Campj'lopus exasperatus 














+ . 


3200-9000 




, , 


+ 








Eucamptodon edentulus. . 
















5000 














Leucophanes squarrosum 
















6000 


+ 












*Sclilotheimia splendida . . 














. • 


6000 














Milium rostratum 
















5000 














Pogonatum macrophyllum 














+ 


6000 














Taxithelium Dubyamim. . 














+ 


50U0 




+ 










*Mniodendron microloma 
















6000 














Korthalsii 














+ 


5000 






+ 








Hypnodendron Eeinwardti 


i . . 












+ 


6000 




+ 










Rhaoopilum spectabile . . 








_ 






+ 


6000 


+ 












Plagiochila tenuis 










.. + 






6000 




. . 


+ 








Chiloseyphus aselliformis 














+ 


5000 














Mastigolryum, sp 
















6000 














Lepidozia Wallichiana . . 










.. + 






6000 














Trichocolea tomentella . . 














•• 


6000 












_• 



C. Upper MorNTAiN Zone (from 6000 to 10,500 feet). 





S ; 


e3 

• S 


83 




^ 




Altitudes on 




eS 




.S 


c 




1 




1 V 

5 O 

5 t: 


s 






41 


Kinabalu. 


a. 


IS 


'o 


3 




J3 2 


w 


O 


«H 




O 


^ 


PI 




Ph 


< 


Ah 


■«• 


;^ 


GO 
















feet. 














Drimys piperita 












+ 


6600-11,000 


+ 


+ 




-^ 






*Ternstroemia Lowii 












-~^ 




9000 














*Adinandra verrucosa 
















8000 














*Saurauja actiiiidiifolia . . . 












^^ 


-^ 


7000 


-^ 


—■ 










8chima brevifolia 






-^ 






-^ 


-^ 


8000-11,000 














*EljEocarpus sericea 












-^ 




8800 














Ilex spicata 












+ 


+ 


7000 












1 


* ■ vaccinii folia 








■^- 




— 


-^ 


8800 


-^ 












Microtropis ramitiora 












. . 


+ 


8000 














*Hhus borneeusis 








+ 
+ 




+ 
4- 




6000-8000 

7800 
9000-13,000 

7600 

10,500 
8000-10,500 

9000 
7000-13,000 

7700 






1 








*Pygeum oocarpum 

*Riibus Lowii 












— — lineatus .... 






*PolTosma Hookeri 






* bracteosa 

Leptospermum recurvum . 












+ 




— 


•• 




z 






*Myrtus flavida 
















5500-7000 
6600 
8800 
8800 














*Tristania bilocularis 












■ 


'~'~ 














*Eugenia kinabaluensis .... 










— • 


















* "Mvrtillns 






■ 






.. 1 














' — 




1 



124 



DE. 0. STAPP OX THE FLOEA OF MOUNT KIN AB ALII. 
C. Upper Mountain Zone {continued). 





O 

3 "So 




.s 

o 

o 




o 


Altitudes on 
Kinabalu. 


at 
S 


■ ^ 

2S 


.2 

C 
o 


.5 

a 

3 


1 
1^ 


i 

a 

a c 

1^ 


f 


q 5 1 


a 


W 6 ^ 


p 




Ch 


< 


Ph 


< 


--Ti 


/3 
















feet. 














*Sonerila crassiuscula 












— 


^ 


9U00 














*}*Iediiiilla stephanostegia .... 








-^ 




^- 




7600 

7600 

7000-13,000 














*Begonia Burbidgpi 












- — - 


-^^ 














*Trachvmene saniculaefolia .... 






. . 


. . 


, . 












"^ 






*Hedyotis pulchella 

* macrostegia 














— 


7000-9000 
8000-12,000 

6600 

7000 


— ' 












*Lucina3a pentacme 














■ ■ 














* nervulosa 














■ — . 














*rrophYHum cypliandrum 












-^ 


■^- 


7800 
7600 
10,500 


























•-^ 


-^ 














* subanurum 




*Psychotria densifolia 












— 


•• 


8000 
7600 














*Lasianthiis euneurus 










'^- 


















* rotundatus 






, , 




-^ 






8300 














Nertera depressa 












+ 


+ 


10,500 


+ 


+ 


+ 


+ 


+ 


+ 


Coprosma crassicaulis 
















10,500 








"■^ 


"^ 




Vaccinium coriacevira. 














-^ 


9000 














buxifolium 1 












-^ 


7500-11,800 














* cordifolium 












, 


, 


8800-9000 














*- pacbydermum 

, sp 

*Diplyco3ia chrysotbrix 

* rufa 








•• 




-^ 




5500-10,500 

8800 














• • 










-^ 


— 


6600 
7500-8000 














*— kinabaluensis 
















10,500-12,000 














ciliolata 








. . 




. , 




6000-8000 














heterophylla (?) 


. . 










+ 


-f 


7600 














* punctulata 


, , 










-^ 


-^ 


7000 














' * cinnamomifolia 








, , 




-^ 


-^ 


8000-9000 














' * MyrtiUus 


, . 


^ , 






. . 


— 


— 


8700 














Khododendron Lowii 








, , 




-^- 


-^ 


8000-10,000 


— 


+ 










■ Brookeanura 














-^ 


7800 














stenopbvllum 












, , 


-^^ 


5000-8000 














rugosum 














-^- 


7000-10,000 














acuminatum 












-~^ 


-^ 


6000-8000 


-^ 












durionifolium 












-^ 


+ 


4000-7000 















verticillatum 














+ 


8000-11,500 














i * cuneifolmm 








-~^ 




. . 


-^ 


7000 


— ^ 












■ ericoides 














-^ 


8000-10,500 


^^ 












Lcucopogon suaveolens 














+ 


7000-13,000 


+ 






— 


— 




Mvrsine capitellata 








+ 


+ 


+ 


+ 


8000-11,500 














*Embelia tortuosa 
















8800 














■ coriacea 


.. 










+ 


+ 


7600-8800 














* minutifolia 












— 




8500 


'-^ 












Ardisia javanica 












+ 


+ 


8800 














Symplocos adenopbylla 












+ 


+ 


5500-7000 














zizvpboides 










. . 


— ■ 




10,500 














* deflexa 








. . 




■ — 




8800 














Alvxia lucida 












+ 


+ 


7500 














Geutiana borneensis 


-,_ 


-,^ 




-^_ 


■ — 


-^ 


^- 


7500 














*EupbTasia borneensis 














i . . 


8000-12,000 








— 







Diehrotricbum asperifolium. . 








, , 




+ 


-t- 


8000-10,500 




+ 










*Didymocarpus areolatus 






, , 










5000-7000 














Cyrtandra Clarkei 












^ 




10,500 














*Strobilantbes galeopsis 


1 


, , 


, , 


, , 








5000-7000 














1 Nepentbes Lowii 


. 








, , 






6000-8000 














villosa 












. . 




7000-9000 














Edwardsiana 


i • 














4000-9000 















DE. O. STAFF OX THE FLORA OF MOUNT KINABALU. 
C. Upper Mountain Zone (continued). 



125 





§ 


B 

<B a. 


i 

9 




a 

c 




d 

01 


Altitudes on 
Kinabalu. 




el 


a 


1 


1^ 










O 

-a 

s 

M 


a 
n 


1 




1. 
o 




0^ 




"3 


^ 

s 
< 


r 


















feet. 










Nepenthes Rajah 
















+ 


4000-7000 














tentaculata 
















-1- 


0000-0000 














Litsaea zeylanica 








. . 


-^ 


■i- 


+ 


+ 


8000 


+ 


+ 










Lindera caesia 










.^. 




+ 


+ 


8000 














*Loranthus sabaeusis 


















9000-11,000 


-^ 












* centiflorus 


















6500 














Daphniphyllum borneense 






— . 


-,_ 


^^ 


-^ 






10,000-11,000 














*Ficus setifiora 



















7000 














* kinabaluensis 










"^ 




,^ 




8000 














ilotleyana 
















+ 


8800 














Elatostemma bulbothrix . . . 


















10,000 














*Quercus Havilandii 
















-^„ 


0000-10,5(10 


+ 












*Castanopsi8 turbinata 
















^^ 


8500 














, sp. 


















8000 


— 












Platj'clinis granditiora 


















10,5ti0 














*Dendrobium kiiiabaluense . . . 


















7000 














*Bulbophylhim catenarium . . . 




. 








• • 


+ 


+ 


6500 














* coriaceum 


















10,500 














1 * altispex 














-^ 




8000 














*Eria grandis 


















10,000-12,000 














*Nephelaphylluin latilabre . . . . 














-^ 




6500 














Calanthe parvitlora 












_ 


+ 




7000 














*CcElogyne papillosa . 


















8000-10,500 














*Habenaria borneensis 


1 . 
















10,000 














*Patersonia borneensis 


1 
























-^ 






*Aletris foliolosa 


1 












+ 


.^^ 


5500-7700 














* rigida 


















8000-11,500 














Stnilax laivis 








+ 






+ 


+ 


7800 














Dianella ensifolia 










+ 


+ 


+ 


+ 


5500-8000 


+ 


+ 


+ 


+ 






*Erioeaiilon Hookerianura . . . . 








-^.„ 










5000-7700 














*Cladiuin samoense 


















5500-11,000 




. . 


+ 








Schoenus melanostachyus . . . . 


















5500-11,000 




. . 




+ 


+ 




Carex hypsophila 














+ 




1 0,000 














filicina 










+ 


+ 


+ 




10,500 














Isachne Kunthiana 






— , 


-^ 




+ 


+ 


-^ 


8000 


-^ 


^- 




-^ 


^- 


-^ 


*Deyeuxia epileuca 


















10,500 








■-- 


-- 




Phyllocladus hypophylla . . . 
















+ 


8000-12,000 




+ 










Dacrydium elatum 








+ 






+ 


+ 


6600 














\ sp . 














^ 




8000-11,000 














Podocarpus cupressina 








+ 






+ 


+ 


6600-11,000 






+ 








Gleichenia circinata 
















+ 


5000-10,500 


. . 


, . 


+ 


+ 


+ 




*Cyatht'a polypoda 


















7000 














* Havilandii 


















10,500 














Hymenophyllum dilatatum . 








+ 






+ 


+ 


5000-7000 


+ 


+ 


+ 




+ 




*Tri(;homanes microchilum . . . 


















7000 














■ rigidura 






+ 


+ 




+ 


+ 


+ 


5000-7000 


+ 


+ 


+ 




+ 


+ 


. apiifoliura 














+ 


+ 


6000 


+ 


+ 


+ 




+ 




Davallia contigua 




! 


. . 




+ 


+ 


+ 


1« ,500 


+ 




+ 








Yeitchii 














+ 


+ 


6000-7000 


+ 


+ 


+ 








alpina 














+ 


+ 


10,500 






-h 








Lindsaya janaesonioides 


















5500-9000 














Pteris incisa 








+ 


+ 


+ 


+ 


+ 


7000 


+ 


+ 




+ 






Asplenium laserpitiifolium . . 








+ 


+ 




+ 


+ 


8000 


+ 


+ 


+ 


+ 






*Kephrodiura gymnopodum . . . . 


















10,500 














Polypodium soridens 
















+ 


8000 


+ 












* Havilandii 


















10.500 














cucullatum 












+ 


+ 


+ 


10,500 


+ 


. 


+ 








blechuoidcs 














+ 


+ 


10,500 




+ 


+ 


+ 







SECOND SERIES. BOTANY, VOL. lY 



126 



DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 
C. Upper Mountain Zone {continued). 



Polypoclium stenophyllum , 
laciniatum 



Tittaria sulcata 

Acrostychmra bicuspe 

Lycopodium casuarinoides 

voliibile 

Cainpylopus exasperatus 
*Sematophyllum panduriforme 
Schlotheimia rubiginosa 



!» 




c 




o 


p 


'^'^ 


nj a, 






pq 


O 



+ 



+ 






w 



+ 



+ 

+ 

+ 



4- 



+ 
+ 



Alritudes on 
Kinabalu. 



feet. 

10,500 

10,500 

10,500 

4000-8000 

8700 

8000 

9000 

6500 
10,500 



+ 



-I- 

+ 



5?i 



+ 
+ 



+ 



+ 



D. Summit Zone (above 10,500 feet). 



*llaiiunculus Lowii . . 

Drimys piperita . . . . 

Eurya reticulata 

ftchima brevifolia . , 

*Ilex revoluta 

*B,ubus Lowii 

Potenlilla leuconota 

Mooniana . . . 

* parvula 



o ^ 



*Stranvaesia integrifolia . •. . . 

Haloragis micraufha 

Leptospermtim recurvum . . 

*Eugonia ampullaria 

*Tracbymene saiiiculaefolia . . 

*Hedyotis maerostegia 

*Coprosma Hookeri 

Vacciniiim buxifolium . . . . 

*Gaultheria borneensis . . . . 

I *Diplycosia kinabuluensis . . 

I Ebododendron verticillatura 

buxifolium 

Leueopogon suaveolens . . . . 

Myrsine capitellata 

* dasypbylla 

*Embelia spiraioides 

*Symplocos buxifolia 

*Gentiana lycopodioides . . . . 
*Havilandia borneensis . . . . 

*Euphrasia borneensis , 

! Drapetcs ericoides 

i *Loranthus sabaensis 

Balanopbora elongata 

j *Dapbnipbyllum borneense . 

*Pilea Jobniana 

1 *Elatostemma litboneuron . 



+ 



m 



+ 



+ 



+ 



+ 
+ 



+ 



+ 



+ 



+ 



+ 



+ 



+ + 



+ 



+ 



+ 

+ 



+ 



+ 



+ 



Altitudes on 
KinabaUi. 



feet. 

11,000-12,000 

G600-l],000 

11,000-13,0<»0 

8000-11,000 

11,000 

9000-13,000 

11,000-13,000 

11,000 

11,000 

11,500-13,000 

12,000 

7000-13,000 

11,500 
7000-13,000 
8000-12,000 

13,000 
7500-11,800 

12,000 

10,500-12,000 

8000-11,500 

12,000 

7000-13,000 

8000-11,500 

11,000 

11,500 

13,000 

11,000 

11,000-11,500 

8000-12,000 

12,000-13,000 

9000-11,000 

11,000 

10,000-11,000 

11,000-12,000 

11,000 



+ 



P4 



^'N 



3-5 



+ 



+ 



+ 



+ 



+ 



+ 



DR. O. STAFF ON THE FLORA OF MOUNT KINABALU. 



127 



U. Summit Zone {confmued). 





c 
o 


A 


rs 

a 

+ 
+ 

+ 
+ 


1 

+ 
+ 

+ 
+ 


i 


i 

St 


1 

^ 

■f 
+ 

+ 


Altitiiti.* oti 
Kinabalu. 


i 

c 
'S. 

a. 


•< 


.5 


i .a 

i c 

a 


' -i 

+ 
-1- 


i 

1^ 


1 


ilyrica javanica 

*Platyclinis grandiflora 

* stachyodes 

♦Biilbophyllum moutensc ... 

*Eria grandis 

j *Aletri8 rigida 

Scirpus inundatus 

1 * Clarkci 




+ 


+ 


+ 
+ 


11, 000-11, 500 

10,5(>O 

11,0(K» 

11,000 

Iit,(K>0-12,0(^>0 

80<.>0 11,500 

11,000 

11,000 

1 1,(KK) 13,000 

55(M»- 11,000 

11,000 

ll.OlK) 

13,000 

13,000 

NM»0-1LVW»0 

0000-1 l.fH»0 

11,000-12,000 

11,(KH> 

11.000 

It, (too 

11,000 
11,500 
11,500 
11,0CK) 
11,500 
11,500 




i 


4- 


1 

! 

j 

i 
i 
j 


Schoenus apogon 

*Cladium samoense 

Carcx rara 








fusiformis 

' Descharapsia flexuosa 

Agrostis canina 


+ 


.. 




+ 


J. 


Pliyllocladus hypophylla . . . 

Podocarpus cuprctisina 

neriifolia : 


4. 


+ 






Lycopodium zeylanicum .... 
Sclago 


+ 




O. 


Wiglitianum 

clavatum 

Dicranura, sp 

Ithacomitrium javanicum .... 
Macromitrium cuspidalum .... 

*Harrisonia alpina 

Udontoschisma Sphagni .... 















EXUMERATIOX OF THE PLANTS COLLECTED ON MOUNT KINABALU. 
WITH DESCRIPTIONS OF NEW SPECIES. 



RAXUNCULACE.E. 

RAXUNCLLrs Lowii, Stapf, in Hook. Ic. PI. t. 2261. Rliizoma brevo, o])liquum vol sub- 
vcrticaliter doscendens, fibrillis radicinis et fibris (foliorum omarcidorum rcsiduis) 
creberrime Tcstitum, Fob'a omnia basalia, i-G, petiolo 1-7 poll, longo siiffuUa, 
rotundata, raajora 1-]|- poll., minora intcrdum vix J poll, longa lat^ique, ad fj 
triloba, lobis brevibus, latis, apiculatis, integris vel 1-2-dentatis, supra pilis flavidis 
adpressis, ut petiolus striguloso-birsuta, subtus i:la!)ra. Pedunculi 1-4, .scaposi, 4-8 
poll, longi, apbylli, adpressc birsuti. Sepala adprcssa, ovato-lanccolata, 2-2f, lin. 
longa, obtusiuscula, margine membranacea, plus minusve hirsuta. Petala obovato- 
oblonga, 4f-5 lin. longa, 1^ lin. lata, aurea, subtus rubro-venia, glabra, basi squamul^. 
parv^ obtus^ nectarium fovente instructa. Achaenia numerosa, receptaculo hirsuto 
insidentia, arete congesta, oblique ovata, 1 lin. longa, subinflata, glabra ; stylus h lin. 
lono-us, rostriformis^ curvatiis, apice involiitus. 
In moist sunny places, from 11,000 to 12,000 feet {Lov: ; Haviland, 1055). 
Sir E. von MueUer described a very similar species from Mount Victoria in New Guinea 

under the name of JR. amerophyllus (Trans. Roy. Soc. Victoria, vol. i. pt. ii. 1). I have 

s2 



128 DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

not seen it, but it seems to differ chiefly in the narrower, comparatively longer, leaves and 
the very dwarf habit. In any case both species are closely allied to, and representatives 
of, a set of Banunctdus-s^ecies which inhabit Australia and Tasmania, and group round 
the multifarious B. lappaceus, Sm. B. Lowii approaches most nearly in habit the variety 
scaijigerus of B. lappaceus (particularly certain forms from the highest parts of the 
Australian Alps and the alpine region of Tasmania), and B. Muelleri, Benth., which, 
however, differs very remarkably in the shape of the beak of the achsenium. So far 
as I can see from Mueller's description, B. amerophyllus seems to come nearer to 
B. lappaceus than the Kinabalu plant does. 

MAGNOLIACE^. 

Drimys piperita, Hook. fil. in Hook. Ic. PI. t. 896; F, von Mueller, Trans. Roy. Soc. 
Victoria, vol. i. pt. ii. 1. 

From 6600 to 11,000 feet (Zoiv ; Haviland, 1175, 1176, 1177, 1178). 

Pistribution : New Guinea to the Philippines and Borneo. 

The plant was found by Beccari in the Island of Amboina, and lie says ('Malesia,' i. 
p. 18) it is probably the same species which he met with on the tops of almost all the 
mountains he visited in that region, and at least no. 3235 of his Sarawak collection 
belongs to it. It was discovered by Sir William Macgregor on Mt. Knutsford in 
New Guinea, and I think there is hardly any doubt that specimens from the Philippines 
( Vidal, no. 2280, from the district of Lepanto, North Luzon, and no. 3768, without 
locality) must also be referred to D. piperita. This species is nearest allied to 
D. dipetala, F. Muell., a native of New South Wales. 

Illicium, sp. Frutex. Ilami juniores striato-angulati. Folia elliptica, 4-5 J poll, 
longa, 2^ poll, lata, obtusa vel subacuminata, coriacea, nervis lateralibus primariis 
7-8, 6-8 lin. distantibus. Pedicelli graciles, stricti, Ij poll, longi. Sepala ovato- 
elliptica, obtusa, 3 lin. longa. Petala similia, latiora, 3-4 lin. longa. Stamina 9, 
vix 2 lin. longa, filamentis oblongis vel ellipticis; antherarum loculi introrsi 
subparalleli. Carpella 8-9 ; ovulum ex ima basi ortum. 
Kinitaki, 4000 feet {Ilaviland, 1272). 

As I have seen only a branch and two detached incomplete flowers I abstain from 
giving this plant a specific name, the more so as the limitation even of the better known 
species from Indo-China is still far from being satisfactory. In any case, the Bornean 
plant certainly is closely allied to the continental species, perhaps most to T. camhodianum, 
Hance. 

MiCHELIA (?), sp. 

Maripari S2)ur (Low). 

There is only one branch witli a flower-bud in the collection. I have not been able to 
identify it with any of the Magnoliacese in the Kew Herbarium. Mr. Burbidge 
also mentions a *' Maynolia " in one of his note-books. He found it in the same locality, 
and describes it as possessing ovate and coriaceous leaves and flowe)\s of 2-4 inches 
diameter, of a yellow colour " inclining to fuscous/' and of the most delicious fragrance. 



DR. 0. STAFF ON THE FLORA OF MOL^NT KINABALU. \'2\) 

It is no doubt the saine plant as Sir Hugh Low's. As the material is IraL^nientaiy, I 
cannot go beyond the suggestiou that it is more likely a species of Michel ia, 

ANONACE^. 

GoNiOTHALAMUS DISPERMUS, Miq. {?), Ann. Mus. Luijd. Bat. ii. 3A; Schelfer, Ann. 
Jard. Bot. Biiit. ii. 25. 

At 3000 Feet {Uaviland, 1313). 

Distribution : Borneo. 

G. dispermus, so far as it is represented by a specimen of Miquel's in the Kcw Herbarium 
and by a specimen collected by Motley (no. 1127) near Banjermassin, resembles very 
much G. malayamis. Hook. fil. & Th., but the flowers are twice or three times as large, 
attaining 2-2| in., and golden yellow, not greenish. The plant from Kinabalu has 
flowers of the same size, shape, and colour; the sepals and the leaves, however, ditter 
slightly. The former are much less hairy, and the outer ones are, except at the base, 
almost glabrous outside. The difference in the nervation of the leaves is perhaps of 
more importance. The nerves are very faint and, in leaves of equal length, less 
numerous, 9-10 against 16-17 in the typical plant, and the transverse venation, which, 
although very faint, can be distinctly seen in the South Bornean specimens, is almost 
invisible in the Kinabalu plant, even when the leaves are held against thi*. bght. 
Besides, the leaves of Ilaviland's plant are more suddenly contraeted into the linear 
blunt acumen. But, taking into consideration the variability of these characters in other 
species which are better known, I doubt whether they are sufficient to separate tlu' 
North and the South Bornean plant specitically, and I the less venture to do so as the 
material at hand consists only of three specimens. 

GoNiOTHALAMUS STEXOPETALUs, Stapf, n. sp. Cortcx ramoruui floriferum griseo-bruneus, 
ramulorum bruneus. Folia petiolo nigricaute supra valde canaliculato robusto 
4-5 lin. longo suffulta, oblonga vel oblongo-lanceolata, 8-10 poll, longa, 2.}-3 poll, 
lata, basi pffine rotundata, abrupte breviterque acuminata, coriacea, utrinque opaca 
glabraque, nervis lateralibus primariis tenuibus, utrinque 13-10, arcuatis et in 
triente exteriore valde prorsus ductis, supra subtusque prominulis, venulis laxe 
anastomosantibus. Flores singuli, e trunco vel e ramis vetustis orti. Pedunculus 
1-,1J poll, longus, nutans, basin versus bracteolis 6-7 parvulis obsitus. Sepala 
basi connata, triangularia, 3 lin. longa, acutiuscula vel subacuminata, glabrata. 
Petala exteriora e basi ovatd longissime caudato-acuminata, 1-1| poll, longa, 2 lin. 
lata, porrecta, glabrescentia vel pilis rufis adpressis, imprimis basin versus minute 
strigillosa, uninervia ; interiora ovata, vix unguiculata, in marginibus albido-, 
cgeterum rufo-puberula, ad 7 lin. longa, 3-3J lin. lata. Stamina numerosa, linearia ; 
connectivum depressum, rhomboideum, minutissime papillosum. Ovaria numerosa, 
o>labra ; stylus breviter rostnformis, incurvus ; stigma truncatum, unilateraliter 
scissum ; ovulum basilare solitarium. Carpella matura globoso-ovata, 5 lin. longa, 
stipite dimidio minore suffulta^ monosperma. 
Penokok, alt. 3000 feet {Ramland, 1217). 



130 DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

G. stenopetalus is a very well-marked species, distingaislied by the long and narrow 
outer petals. It approaches most nearly G. macrophyllus. Hook. fil. & Th., a species 
indigenous in Sumatra and Malacca. 

GoxiOTHALAMrs KOSEUS, Stapf, n. sp. Cortex ut in G. steMopetalo. Folia quoque consi- 
millima, imprimis quoad nervationem, sed majora, pro ratione angustiora (11^ poll, 
longa, 2\ poll, lata), caeterum valde variabilia, in ramulis novellis interdum vix 4 poll, 
longa. PedunculuslJ-'lf poll, longus, robustior. Sepala basi connata, late triangu- 
laria, 3-4 lin. longa, acuta, glabrata. Petala exteriora rosea, e basi late unguiculatd 
ovato-lanceolata, apicem versus longe attenuata, ad 2^ poll, longa, 6-7 lin. lata, por- 
recta, nninervia, fere glabra, crassiuscula; interiora ebasi indistincte unguiculata ovata, 
ad 10 lin. longa, ad 4-5 lin. lata, acuta, in marginibus parce albido-puberula, caeterum 
fere glabra, crassa. Stamina numerosa, linearia ; connectivum depresso-capitatum, 
minutissime papillosum ; antherarum loculi distantes. Ovaria nuinerosa, ovato- 
oblonga, aureo-hirsuta, in stylum 2~^-plo longiorem, angulo fere recto curvatum 
abeuntia ; stigma truncatum, unilateraliter fissum ; ovulum basilare, solitarium. 
At 6000 feet {Haviland, 1312). 

This plant is so very similar to G. stenopetalus in the vegetative parts that I should not 
hesitate to refer it to that species but for the remarkably different flowers. These are twice 
or three times as large, and have so different a shape that I cannot think this could be due 
to a more advanced state ; besides, they are much thicker and more fleshy. 

MiTnEPHOKA Maingayi, Hook. fil. & Thorns, in Hook. fil. M. Brit. Ind. i. 77 ; King, 
Mat. ri. Malay Pen. no. iv. 86. 

At 3000 feet, Dahombang Iliver {Haviland, 1311). 

Distribution : From Malacca to South Burma, Java, and Borneo. 

I refer this plant with some reluctance to Jf. Maingayi. The leaves are more 
membranaceous and more glabrous than in any specimen I have seen of M. Maingayi. 
On the other hand, the floral structure is the sanie, and the same venation of the 
leaves may be found also in some of the specimens of M. Maiyigayi of the Malay 
Peninsula. Scheffer described a plant under the name of M. glahra (Ann. Jard. Bot. 
Buit. ii. 20) from Selaboun in Borneo, which evidently is very similar to that from Kina- 
balu ; but there is nothing in the description that would not apply also to M. Maingayi, 
at least in the broad sense in which it is understood by Dr. King. 

Melodobum KiNABALrENSE, Stapf, n. sp. Alte scandens, ramis pendulis. Barauli 
novelli pilis strictis, dense breviterque fulv o-tomentelli, demum glabrati, nigricantes. 
Folia petiolo 4-5 lin. longo, supra vix sulcato, fulvo-tomentello, deinde glabro 
nigricante suffulta, oblongo-elliptica vel oblonga, 3i~4 poll, longa, Ij-l^ poll, lata, 
basi rotundata, ohtusa, acuta vel acuminata, tenuiter coriacea, supra brunea, in costa 
strigillosa, cjjeterum pilis minutis rigidis demum fere detritis ohsita, sul)tus magis 
olivaceo-viridia, ubique pilis rigidis adpressis minutis strigilloso-puberula, nervis 
lateralibus 10-13, sul) angulo 40°-45' ortis, apice valde prorsus curvatis, vix vero sub 
margine anastomosantibus. Flores axillares, singuli. Pedunculus primo apice 



DR. O. STAFF ON THE FLORA. OF MOUxNT KINABALU. 131 

nutans, sub anthesi strictus, 1 poU, longus, gracilis, fulvo-tomentellus, basi atquc ad 
medium vel supra vel basi tautum bractcolatus. Sepala vix connata, triangularia, 
1-lJ lin. longa, extus fulvo-tomentelln. Potala oMeriom ovata, 5-0 lin. longa, a-t 
lin. lata, obtusluscula crassa, margiuibus iutus deprcssis, fere 1 lin. latis, extus pilis 
adpressis densissime fulvo-subsericea, intus miuute griseo-puberula ; inieriora 
ovata, 2 lin. longa, acuta, valde convexa, extus griseo-fulvo-tomentella, intus 
glabra. Stamina numerosa ; filamenta brevissima ; antbemrum loculi contigui ; 
eonnectivum glabrum, ovatum, apice iuflcxum. Ovaria numerosa, oblique ovoideii, 
fulvo-birsuta ; stylus brevis ; stigma pilis superatum ; ovula t in serie unictt. 

Penokok, alt. 3000 feet {Haviland, 1310). 

This species is closely allied to Melodorum fuUjens, Ilook. iil. The indumentum is 
exactly of the same character as in M.falgens, but less shining on the outer petals. 

VIOLACE^. 

Viola SERPENS, Wall, in Eoxb. Fl, Ind. ed. "Wall. ii. 449, and in DC. I'rodr. i. lod. 
Hook. fill. & Thoms. in Hook. fil. Fl. Brit. Ind. i. 181. 

On the Dahombang River, alt. 3000 feet [Ilaciland, 1275). 

Distribution : Throughout tropical Asia, except the Philippines and East ^lalaya; 
temperate Himalaya. 

According to Dr. Haviland the flowers are slightly scented and white, and tlic lip 
streaked with violet. The leaves are perfectly glabrous, even in a young state, and the 
stipules fimbriate. The valves of the mature capsule are about 4 lin. long; besides, 
there is a capsule on Dr. Haviland's specimens which is globose and 2J lin. long. It 
developed evidently from a cleistogamic flower. I think, if we take V. serpens in the 
broad sense of the ' Flora of British India,' we should include also V. distans, WaU. 
The characters taken from the fruit are, in my opinion, reducible to heterocarpy, the 
capsules derived from normal flowers being always more or less oblong, those from 
cleistogamic flowers globose. A similar heteromorphy seems to exist with regard to the 
shape of the stigma, as Hook. fil. & Thoms. have suggested. But, apart from this, it is not 
unlikely that several species of close affinity are included in V. serpens, taken in this 
broad sense. Yet it seems to me quite hopeless to decide that question from dry 
material only. In any case, it is evident that one species — if not several nearly 
related species — is spread from the temperate region of the Himalaya and from South 
China all over the mountain-region of the Western and Eastern Peninsula, of Ceylon 
and Malaya, and that this species, or this group of species, is most closely allied 
to forms which are typically Boreal. 

POLYGALACE.E. 

PoLYGALA VENENOSA, Juss. in Poir. Diet. V. l')3. 

On the Tampassuk River near Kiau, from 2000 to 3100 feet {Low, Haviland, 1319). 
Distribution : West Malaya to the Philippines. 



132 DE. O. STAPr ON THE FLOEA OF MOUNT KINABALU. 

HYPEEICINE^. 

Hypericum mutiltjm, Linne, Spec. Plant, ed. I, 717. — Confer Maximowicz, in Bull. Ac. 
Sc. St. Petersb. xi. 171. 

In buffalo pastures at Kiau, alt. 3000 feet {Saviland, 1331). 

Distribution : Japan ; Central and South China ; India and Ceylon ; Sumatra (Beccari, 
867), Java and Borneo ; New Zealand and Tasmania ; Madagascar ; North and South 
America. 

A weed, probably of American origin. 

Ckatoxylon polyaxthum, Korth. Verb. Nat. Gesch., Bot. 175, t. xxxvi. 
Tawaran River, alt. 1500 feet {Eaviland, 1386). 
Distribution : Malaya to South Burma, South China, and the Philippines. 

GUTTIFER^. 

Garcinia Havilandii, Stapf, n. sp. Ramuli subtetragoni, brunnei, lenticellis minutis 
striolati. Polia petiolo circa 4 lin. longo suffulta, elliptica vel ovato-elliptica, 4|-6 
poll, longa, 2-2^ poll, lata, basi breviter lateque cuneata, apice abrupte in acumen 
obtusum lineare ^ poll, longum contracta, supra olivaceo- vel obscure viridia, subtus 
pallidiora, nervis niimerosissimis teneris^ titrmque distinctis, subhoj^izontalibns, 
J-f lin. distaniibus, Flores mascidi in cymulis paucifloris axillaribus dispositi; 
pedunculi 1-1^ lin. longi; pedicelli circa 2 lin. longi, bracteis minutis suffulti. 
Sepala 4, decussata, exteriora 2 rotundato-ovata, f lin. longa ; interiora rotunda, 
valde convexa, duplo longiora. Petala 4, imbricata, rotunda, IJ lin. longa, alba. 
Stamina numerosissima, in phalangis 4 late rotundatis basi connatis ; anthera^ 
subsessiles, subglobosae, coacervata?, rima apieali vel (in inferis) longitudinali 
dehiscentes. Ovarium rudimentarium pileiforme, stipite phalangas yequante vel 
superante, disco integro lato, circa stylum rudimentarium brevissimum leviter 
impressum. 
On the Tanipassuk River, near Kiau {HarAland^ 1379). 

The floral structure is almost the same as in G. mergiiiensis, Wight, but the leaves 
are much larger and show a different venation. As the Kew Herbarium possesses all 
the types of Pierre's new species of Garcinia from Borneo, and nearly all the other 
Malayan species whicli possibly might come under consideration, and as I am not able to 
identify the plant from Kiau with any one of them, I do not hesitate to describe it as a 
new species, though the material is not perfect. It belongs to the section Discostigma^ 
Pierre, a group of distinctly Malayan origin with a feeble outlier in Ceylon. 

TERNSTRCEMIACE^. 

Ternstecemia Lowii, stapf, n. sp Erutex dioicus. Ramuli crassi, cortice grisco, 
exsiccato valde ritgoso et irregulariter sulcato. Eolia subvcrticillata, petiolo J poll, 
longo crasso suffulta, oblonga vel obovato-oblonga, 2-5 poll, longa, 1-2 poll, lata, 
basi cuneata, rotundato-obtusa vel subemarginata, valde coriacea, rugulosa, supra 



DE. O. STAFF OX THE FLORA OF MOUNT KIN.ABALU. 133 

castanea, ;plerumque nitid'mima, nervls valde obliquh ind is/ if tcf is utnn({ue 5-7, subtus 
oj)aca, i)a\lidiora, etiervia. Mas: Pedicclli (juveniles) :\ lin. longi, crassi, rccurvi, 
sub ipso flore bibmcteati ; bractew ovatae, carinata?, crassi, 1 lin. loni^nr. Sepala 
rotundata, crasse coriacea, valde convexa, interiora 3 lin. loni^a. Petala obovato- 
rotundata, tenuiter membranaceo-marginata, vix \ lin. loni^a. Stamina numerosa, 
polyadelphia ; antherae 1-1 J lin. longic, linoari-obUmgje, filament oruni partem 
liberam sequantes yel breviores ; connectivum minutissime apiculatum vel baud 
productum, obtusum. Femina ; Fructus pedicello incurvo 5-6 lin. longo suffultus, 
basi sepalis persistentibus valde coriaceis majoribus vix 3 lin. longis circumdatus, 
globosus, coriaceus, 6 lin, dimetiens; loculi 2; fiemina i-f) in iitroque loculo, 
oblonga, triquetra, 2J lin. longa. 
Maripari Spur, alt. 5000 feet [Ilaviland, 1262 6 ), alt. 9000 feet {Uaviland, 1129 6 ; 
Low, $ ). 

Ternstrcemia Loioii is nearest allied to T. hancana, ^liq. [Adinandra bancanay King), 
and T. coriacea, Scbeffer, but distinguished from botli by the generally very shiny veinlcss 
leaves, which in a dry state often look as if they were varnished ; from the former also by 
the more terete and comparatively slender branches and the recurved and more sl(»nder 
pedicels ; and from the latter by the smaller and very thick sepals, by more globose and 
smaller berries and more numerous and smaller seeds Avith acute angles. 

Adinandra excelsa, Korth. Verb. Nat. Gesch. Bot. 109, t. xviii. 

On the Tampassuk Hiver near Kiau, alt. 2000 feet (Ilaviland, 1373). 

Distribution: Borneo. 

A. excelsa comes very near to A. acuminata, Korth., particularly as this is represented 
by Korthals's type ; but it differs, in my opinion, sufficiently in the venation of the leaves, 
especially in tlie very characteristic marginal arching of the nerves. 

Adinandka verrucosa, Stajif, in Hook. Ic. PI. t. 2266. Arbor parva. Piami crassi, 
nigrescentes, demum pallidiores, glaberrimi, mox lenticellis pallidis prominentibus 
verrucosi. Folia subdisticha, petiole crasso supra acute sulcato 3-4 lin. longo 
suffulta, elliptica, 5-6 poll, longa, 2^-3 poll, lata, obtusa, emarginata, saepe sub 
apice paulo contracta, valde coriacea, gla])errima, costa supra leviter sulcata subtus 
valde prominente et bisulcat4, nervis lateralibus circa 15 inconspicuis. Pedicelli 
axillares, fasciculati 3 vel 2 vel solitarii, crassi, rccurvi, 3-i lin. longi, basi bracteis 
ovatis carinatis adpresse pilosis instructi, apice })racteolas 2 Kite rotundato-ovatas 
margine minute pubescente excepto glabras gerentes. Calyx bracteolis arete 
approximatus, sepalis rotundis 3-4 lin. latis coriaceis membranaceo-marginalis, 
maro-ine puberulo-ciliato excepto glabris. Petala rosea, e basi breviter lateque 
uncuiculata late eliiptica, 7 lin. longa, 1-5 lin. lata. Stamina numerosa, basi 
connata ; filamenta linearia ; antherae lineari-lanceolatse, in dorso argenteo-sericeae 
apiculatse. Ovarium glabrnm, ovato-globosum, sensim in stylum integrum, 6-7 lin. 
lono-um abiens. Fructus circa 4 lin. dimetiens. Semina subreniformia, nitida, 
castanea, punctulata, ^-f lin. longa. 
At 8000 feet (Raviland, 1101). 

SECOND SERIES. — BOTANY, VOL. IV. T 



13i DR. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

A very well-marked species, distinguished by the verrucose branches, the large thickly 
coriaceous leaves, the short thick pedicels, and the large flowers. 

EuRYA RETICULATA, Eorth. Yerh. nat. Gesch. Bot. 118. 

Prom 11,000 to 13,000 feet (Low; Haviland, 1067). 

Distribution : High mountains of the Malay Archipelago from Sumatra to Borneo. 

The Kinabalu plant agrees perfectly with the type specimens of L. reticulata and 
E. coneocarpa^ Korth., which I find are exactly the same. The stems of Haviland's 
specimens are covered with a thick coat of Capnoclmm elongatum^ Berk, et Desm. 

Saurauja terox, Korth. Yerh. nat. Gesch. Bot. 132, t. xix. 

At 3000 feet {Haviland, 1333). 

Distribution : Malaya, from Sumatra to the Philippines (?). 

I refer to this species also Motley, 332, from Banjermassin, Beccari, 1021, from 
Sarawak, 95 and 131 from Mt. Singalang in Sumatra, and a specimen of Korthals, 
named " Saurmija siyigalensis, Korth.," and, though with some reluctance, a specimen of 
Cuming, 1712, from Samar in the Philippines. The specimens from Mt. Singalang dijflper 
but slightly in the more copious indumentum of the leaves. S.ferox is closely allied to 
aS. Beinicarcltiana, Bl., from Java, and S. hirsuta, Bl., from Celebes, and it is very probably 
identical with S. tewensis, Korth., and perhaps also with S. sakoembangensis, Korth., 
both from South Borneo. S. Reinwardtiana differs chiefly in the larger bracts, which 
are very much reduced in S.ferox, a character of apparently little importance, and by 
the more delicate venation, which, however, is of almost exactly the same structure. If 
S.ferox and S. Beiuwardtiana should indeed prove to be only forms or varieties of one 
species, the name S. Beiuwardtiana would stand as the older. A type specimen of 
S. tewensis, Korth., for which I am indebted to Dr. Boerlage of Leyden, agrees almost 
perfectly with the specimen from Kinabalu so far as the foliage and the branches are 
concerned. But the only imperfect flower which is still attached to the branch is 
smaller, and the calyx clothed with a more paleaceous covering. 

Saurai'JA amcexa, Stapf, n. sp. Arbor parva. Bamuli pallidi, fusco-strigillosi, et 

praeterea squamulis -lanceolato-subulatis adpressis conspersi. Polia petiolo indu- 

mento consimili vestito 10-14 lin. longo suffulta, obovato-elliptica, 5-6 poll, longa, 

2:|-3 poll, lata, basi acuta, breviter acuminata, spinuloso-serrulata, supra setulis rufis 

adpressis strigillosa, obscure brunnea, subopaca, subtusin nervis strigillosa, in venulis 

hirsutiuscula, pallida ; nervis lateralibus primariis angulo 45° ortis, utrinque 11-12, 

4^-6 lin. inter so distantibus, marginem versus valde prorsus ductis, nervis secun- 

dariis tertiariisque supra indistinctis, infra distinctis, tenuibus, nonnullis inaequalibus 

e costa angulo 60°-80°, rarius 90° ortis, leviter prorsus flexis in media via ad 

marginem vel prius cum venulis e nervis primariis ortis anastomosantibus, venulis 

prope marginem subrectis, vix 1 lin. inter se distantibus. Inflorescentia axillaris, 

cymosa, bracttata. Pedunculus petiolum aequans vel brevior, ut pedicelli subpatule 

vel adpresse strigosus. Bracteae ovatae vel lanceolata), majores ad J poll. 

longae, structura indumentoque foliorum. Pedicelli sub anthesi 3-4 lin. longi, 



DR. O. STAFF OX THE FLORA OF MOUNT KIXAIULU. 1 3r> 

apice saepe incurvi. Sepala basi connata, ina^qualia, exteriora 3 rotundato-ovata, 
6 liu. longa, 4 lin. lata, obtusa vel subacuta, extus doisissime setia hivsutulis 
incur vo-jKitulis vest it a, iiitus glabra, iuteriom 2 subpetaloidea, oblonga, 5 lin. longa, 
2 lin. lata. Petala alba, ex ungue circa 5 lin. longo 3^ lin. lato cordata ; lamina 
5J lin. longa, 6-7 lin. lata, sinu angnsto. Stamina nuiiicrosa, basi connata; 
filamentoram pars libera antberis vix 1 lin. longis brcvior. Ovarium glabrum, 
hemispbgericum, triloculare ; styli filiformes 3. 
Near Kiau, alt. 2500 feet {Haviland, 13G1). 

Saurauja amoena belongs to the group of S.ferox, but its flowers are larger than in 
any one of the allied species. The venation of the leaves is very similar to that of .V. scti- 
gera, Kortb., and particularly a specimen collected by Beccari in Sumatra, at Ajer Mantjoer 
in Padang, 715; but these dilfer in the size as well as in the indumentum of the 
flowers. 

Saurauja actixidiifolia, Stapf, n. sp. Hamuli squaw ulis lancrolatis primo dense 
f-estiti, mox vero glabrati, jiallidi. Folia petiolo indumento consimili vestito 8-9 
lin. longo suffulta, elliptica vel subovata, 3-1^ poll, longa, lJ-2 poll, lata, Imsi 
subrotunda vel vix acuta, acuminata, setuloso-serrulata, supra setulis l)rcvissimis 
aspersa vel glabrata, subopaca, obscure brunea, subtus in costA, squamulis its pelioli 
similibus, in nereis squamulis vel setulis minutis adpressis laxe vesiita, pallida, nervis 
lateralibus angulo 45° ortis subrectis, sub ipso margine prorsus curvatis, utrinque 10- 
14, 3-3J lin. inter se distantibus, venulis transversis subtus ob colorem obscurum valde 
conspicuis, vix 1 lin. a sese remot/w, numerosis, angulo 5(f--60 ortis, subsigmatoideia. 
Flores axillares, solitarii. Pedicelli graciles, 9-12 lin. longi, laxe squamulosi, sub 
medio bracteold ovatd muniti. Sepala ima basi connata, insequalia, exteriora 3 
rotundato-ovata, 4 lin. longa, 3 lin. lata, acuta vel obtusiuscula, extus paleis 
linear i-subulat is Jursutulis adpressis laxe vestita, intus glabra, interiora 2 angus- 
tiora, oblonga, obtusa, extus paleis perpaucis obsila. Petala et stamina ignota. 
Ovarium (semimaturum) subglobosum, sepalis persistentibus superatum ; styli 5 
filiformes. 
At 7000 feet (Havihmd, 1232). 
The leaves are very like those of Actinidia strigosa. Hook. fil. & Th., whence the 

specific name. A very distinct species of the group of S. ReinicaTdtiana, Blume, well 

characterized by the paleaceous indumentum, the delicate but conspicuous venation, and 

the very much reduced inflorescence. 

Schima breyifolia, Baillon, Hist, des Plantes, iv. 254 (adnot.) ; Stapf, in Hook. Ic. PI. 
t. 2264. — Syn. Gordonia brevifoUa, Hook. fil. in Trans. Linn. Soc. xxiii. 162. 
From 8000 to 11,000 feet {Loic ; Haviland, 1126, 1127). 

Sir Joseph Hooker has pointed out that " the capitate central receptacle of the 
capsule agrees with Plume's character of Eeinwardt's genus Schima J' S. brevifoUa is 
allied to ^S*. Noronhce, Reinw., but it may easily be recognized by the stunted growth, the 
small and crowded leaves, and the larger flowers. 

T 2 



136 DK. O. STAFF ON THE FLOEA OP MOUNT KINABALU 

MALVACE^. 

Hibiscus Abelmoschus, Liune, Spec. Plant, ed. I. 696. 

In rice-fields at Kiau, among long grass and reeds (JBurbidge). 

Distribution : Tropical Asia and Africa ; cultivated also in Tropical America. 

From a fine coloured sketch by Mr. Burbidge. 

STEHCULIACE^. 
Sterculia transltjcens, Stapf, n. sp. Eamuli pube rigidula patula et insuper pilis 
longioribus sparsis vestiti, fulvo-brunei. Eolia petiolo 6-10 lin. longo, indumento 
ramulorum consimili vestito, sub apice subito incrassato suffulta, oblonga vel 
oblongo-lanceolata, 3-5 poll, longa, f-l^- poll, lata, basi acuta, acuminata, Integra, 
utrinque glaberrima, nitidula vel supra tandem opaca, membranacea, translucida, 
nervis lateralibus utrinque 7-8 valde prorsus ductis, subtus prominulis. Hacemi in 
axillis foliorum superiorum, 3-4 poll, longi, graciles, pilis patulis rutis hirsuti. 
Elores 3-5, fasciculati in axillis bractearum lineari-lanceolatarum. Pedicelli 3-1^ 
lin. longi, hirsuti. Calyx oblongo-tuhulosiis, patule hirsutus ; tubus circa 3 lin. 
longus, \\ lin. latus ; lobi lanceolato-subulati, conniventes, 2 lin. longi. Antherae 
vix 10 in apice colnmnae IJ lin. longge arete congestse. 
At 5000 feet {Samland, 1196). 

S. translucens belongs to the section Eu-Sterculia and is nearest allied to S. gracilis^ 
Korth., >S'. rubiginosa, Vent., >S'. eusifolia^ Mast., and still more nearly to an unnamed 
species which was collected at Penang by G. Curtis (no. 526). But the calyx-tube is longer 
in S. translucens than in any one of them, and almost of exactly the same shape as in an 
undescribed species from. Sarawak {Beccari, 1057), which, however, differs totally in tlie 
leaves. 

TILIACE^. 

TmrMFETTA RHOMBOiDEA, Jacq. Select. Am. 147, t. xc. 

On the banks of the Kadamaian Hiver, alt. 2500 feet {Haviland, 1354). 

Distribution : Throughout the Tropics of both hemispheres, particularly in Asia and 
Africa, generally as a weed. 

Triumeetta suefruticosa, Blume, Bijdr. 113. 

On the Kadamaian Biver, alt. 2500 feet [Haviland, 1348). 

Distribution : Malay Archipelago, Solomon Islands, Louisiades (introduced ?). 

This plant was originally described from specimens collected in Java. Later on it 
was found by Ptiedel in Timor-laut and by B. B. Coniins in the Solomon Islands, and 
there are also specimens from the Louisiade Archipelago (Voy. of ' Battlesnake,' 467) in 
the Kew Herbarium. It is very closely allied to T. pilosa. Both, from which it differs in 
the very scanty tomentum of the leaves, which is limited to a few single hairs on the nerves 
beneath and to single or poorly fasciculated hairs above, and in the longer peduncles. 
The bristles of the capsule are either glabrous or ciliate. T. subhirtella, Miq., from 
Java, and T. graveolens, Blume, from Java, and according to Miqucl from Timor, belono- 
evidently also to T. svffruticosa. The specimens from Timor, collected by Biedel, have 



DE. O. STAPF ON THE FLOE A OF MOUNT KINABALU. 1^7 

partly glabrous fruit-bristles, as they are described to be in Triiwf/etla graveolens; 
thus at least I understand tlie term " nudis " in Blume's diairnosis. 

El^ocarpus sericeus, Stapf, n. sp. (§ Acronodia.) Prutex. llanii glabri, juniores 
superne dens'issbnc foliati, vetustiores cortice obscure cinereo. Folia obovato- 
lanccolata, cuneatim in petiolum 2-3 lin. longuni attenuata, 1-U poll. loiii,'a, h \m\\. 
lata, obtusa, obtuse crenata, setis deciduis crenis impositis, coriacea, supra priwo 
adpresse albo-seiHcea, demum plus niinusve glabrat<i, sitbtus tnviissinw dctm'qur 
subaureo-sericea, indumento tandem albescent^?, nervis lateralibus utrinque 6-8 subtus 
tantum distinctis, reticulatione tenui prominuh\. Stipula^ subulato-lanceolata?, 2 3 
lin. longae. Eacemi 1-li poll. lon<^i, paucifiori; rbaehis subaureo-sericea; pedicelli 
2 lin. longi; bractea^ subulata). Sepala l- ovata, acuta, sericeo-velutina, 2 lin. loni,'a. 
Petala i oblonga, basi paulo constricta, apice dentibus 7-8 breviter linearibus. 
Stamina 8, effeta (?) ; filamenta pilosa; antberoe vix 1 lin. longa;, unilateraliter in 
mucronem brevem pubescenteni productaj. Ovarium ovoid(K)-i;lobosum, 2-loculare, 
dense rubro-f ulvo-sericeum ; stylus brevis ; ovula 4 in utroque loculo, biseriata. 

At 8800 feet (Uaviland, 1111). 

Allied to K jyunctalus. King, but very distinct in the stunted lii.bit and the line silky 
tomentum. 

GEllAMACEiE. 

Impatiens plattpetala, Lindl. Bot. Eeg. 1846, t. 08. — Syn. /. latifolia, Miq. 111. 
Fl. Arch. Ind. 93, non L. 

At 5000 feet {Uaviland, 1235). (This is also the Impatiens mentioned by Burbidge 
in ' The Land of the Sun,' p. 286, as a sketch draA\Ti by Mr. Burbidge on the spot shows.) 

Distribution : From Java to Borneo, and probably to New Guinea. 

AYitli this species is very probably identical T. borneensis, Miq, 111. Fl. Arch. Ind. 
96, described from specimens collected by Korthals in South-east Borneo. I have not 
seen the type, but a specimen gathered by Motley in tbe district of Banjermassin 
agrees perfectly with the description, and at the same time with the specimens, of 
I. platypetala from Java. I suspect also that /. cclebica, Miq. 111. Fl. Arch. Ind. 91, is 
a raere sjnonym. of I. plati/j^etala. Miquel does not anywhere mention I. plati/petala, 
Lindl., but there is no doubt that he means the sjime plant by his /. latifolia. This, 
however, is, in my opinion, sufficiently distinct from the true Linnean plant, which is 
found in Ceylon and the southern part of the Western Peninsula. 

RUTACEiE. 

EvoDiA tenuistyla, stapf, n. sp. Frutex scandens. Ramuli minute puberuli, glabres- 
centes, teretes. Folia trifoliolata, petiolo 2|-3 poll, longo minute puberulo supra 
canaliculato suffalta ; foliola insequalia, petiolulis 2 lin. longis suffulta, terminale 
majus, oblongum, circiter 6 poll, longum, 2-2^ poll, latum, basi acutum, anguste 
acuminatum, maro-ine leviter undulatum, tenuiter coriaceum, supra glabrum 
nitidum, subtus pallidum, opacum, in costa et in nervis puberulum, nervis lateralibus 
16-18, venularum reticulatione tenuissima, venis transversis vix uUis; foliola 



138 DE. 0. STAFF ON THE FLOE A OF MOUNT KINABALU. 

lateralia consimillima, circiter 4 poll, longa, 1^ poll. lata. Panicula axillaris, 
thyrsoidea, petiolo paulo Jongior, puberula, minute bracteata, floribus in ramulis 
ultimi ordinis congestis, pedicellis -J-l lin. longis. Sepala late ovata, ^ lin. longa, 
extus puberula. Petala oblonga, vix 1 lin. longa, acuminata. Staoiina valde 
exserta. Ovarium albo-hirsutulum, quadrilobum ; stylus gracilis, apice vix 
incrassafiis, exsertus. 

At Koung on tlie Tarapassuk River, alt. 2000 feet {Haviland, 1376). 

Bvodia tenuistyla has very much in common with E. Hoxhurgliiana, Benth., but it 
differs in the somewhat larger flowers and the slender, not capitate, stigma. 

EvoDiA SUBTJNIFOLIOLATA, Stapf, n. sp. Erutcx. Hamuli pallidi, teretes, tomentelli. 
Eolia opposita, omnia (?) unifoliolata, petiolo circa 5 lin. longo, dense tomentello, 
subterete, apice subarticulato et interdum plus minusve torto suffulta, foliolum 
elliptico- vel obovato-oblongum, basi •kXQvXwva, abrupt e acuminatum, tenuiter coriaceum, 
supra glaucum, subnitens, glaberrimum, infra in gemma dense tomentellum, costd 
exceptd, mox glaJ)ratum., nervis lateralibus utrinque 8-10, l-lj lin. sub margine 
anastomosantibus, utrinque prominulis, venularum reticulatione laxa. Inflorescentise 
axillares, tomentellae, pedunculo 1-1^ poll, longo suffultae, breves, pauciramulosas. 
Elores ignoti. Carpella 1-4, a basi libera, oblique obovoidea, griseo-viridia, 
tomentella, 4 lin. longa, plane aperta, Q-Q\ lin. lata. Semina fere semper solitaria, 
atra, nitida, 2-2J lin. dimetientia, funiculo crasso, vix 1 Jin. longo. 
At 6000 feet {Haviland, 1193). 

The leaflets somewhat resemble those of E. triphylla, EC, but they are covered 
underneath with a dense soft tomentum in a young state, which persists on the mid-rib, 
and they usually terminate with a short narrow acumen. In the specimens I have before 
me the leaves are all unifoliolate, without a trace of lateral leaflets ; but Dr. Haviland 
states on the label " leaves trifoliolate," This may, perhaps, refer to the lower ones. The 
capsules are much larger than those of E. triphylla, and rival those of E. fraxinifolia. 
Hook. til. As to the affinity, it is difficult to form a precise idea in tlie absence of 
flowers. The plant resembles a certain undescribed species from the Island of Panay 
( Vidal, 2275) so far as the leaves are concerned, and it might be a true Melicope. 

MELIACE^. 

Eysoxylon catilifloexm, Hiern, in Hook. fil. & Thoms. El. Brit. Ind. i. 549. 

, var. TOMENTELLA, Stapf. Eolia subtus in nervis tomentella, et prasterea tota planta 

indumento vestita densiore quam in speciminibus in peninsula Malayana collectis 
observatur. 
At the Penokok, alt. 3200 feet {Haviland, 1318). 

Distribution (of species) : Malay Peninsula, from Singapore {Ilidlet, 832) to Penang 
[C. Curtis, 204); Borneo. 

I am not able to find any characters of specific distinction between the Peninsular and 
the Bornean plant. The floral structure is exactly the same in both, except that tlie flowers 
are a little larger, if taken on an average, in the specimens from Singapore than in those 



DE. O. STAFF 0.\ THE FLORX OF MOUNT KIX.VBALU. 131) 

• 

from Kinabalu. The leaves of Maingay's, Ilullet's, and Curtis's specimens ore generally 
glabrous ; but in one of Maiiigay's plants occur a few scattered luiirs on the nerves 
underneath, exactly like those which compose the denser tomentum in th(* Kiiial)ahi form. 

OLACINE^E. 

GoMPHAKDRA LYSIPETALA, Stapf, u. sp. Hamuli 7iovelli adpressissinie minutcqiu' 
puberuli, demum glabmti. Folia petiolo supra tenuissime canaliculato 3J-4 lin. 
longo suffulta, oblongo-lanceolata, 2^-3 poll, longa, |~I poll, lata, basi acuta, 
obtusiuscule subacuminata, coriacea, supra glaberrima, opaca, subtus in gemmA 
dense et minute fulvo-puberula, mox glabrata, nervis lateralibus tenuibus utrinqiie 
3-4. Cyma3 axillares j^edunculo petiolum a^'quante sulTultit?, 5-15-flora^ minutissime 
puberulae vel glabrcscentes ; flores subsessiles, congesti. Calyx breviter cupularis, 
vix J lin. longus, brevissime dentatus, puberulus. Petala valvata, jarmo cohcerentla, 
detnttm plane soluta,patentia, oblonga, l:^lin. longa, glaberrima. Filamenta clavata, 
antice tantuni sub antherls bremter denscque pi/is claiuculatis obsifa, cwterniii 
glaberrima. Antherui sub apice affixiJC, pendului ; loculi paralleli vel snbparalh'li. 
Ovarium rudhnentarium leviter asymmetricum, apice parce ])uberulum. 
At 6000 feet (mwiland, 1207). 
Distinguished by the petals, which finally become quite free and spreading, and l)y 

the peculiar arrangement of the hairs on the filaments. Otherwise tliere is little diirerence 

if compared with narrow-leaved forms of G. axillaris, Wall., G. coriacea, Wight, or 

G. polymorpha, Wight. 

GoMPHAXDKA- PRASINA (1), Beccari, Males, i. 110. 

On the Kadaraaian River, at 4500 feet {Raviland, 12 13). 

Distribution : Malaya, from Malacca and Sumatra to Borneo ; New Guinea. 

Dr. Haviland's specimen is a female with fruits. These are very like those of the 
Malacca plant, called G. affinis by M. Masters in Hook. fil. Fl. Brit. Ind. i. 586, but they 
are narrower and more strongly ribbed. The habit of the plant also differs, the Kinabalu 
plant having more slender branches and smaller leaves. G.Jacanica, Valeton, comes so 
far perhaps nearer, but the shape of the fruit deviates more. The probability, however, is 
that G.Javanica and G. 2)rasina (including G. affinis) belong to one very variable species, 
the oldest name of which would seem to be G. prasina, Becc. 

ILICINE.E. 

Ilex spicata, Blume, Bijdr. 1149. 

At 7000 feet {Eaviland, 1201). 

Distribution: Malaya, from Malacca and Sumatra {Beccari, 130) to Borneo (Beccari, 
3223, 3273). 

Ilex revoll-ta, Stapf, in Hook. Ic. PI. t. 2263. Frutex. Kami stricti, dense foliati, subte- 
retes, tomento minutissimo induti, primo canescentes, mox nigrescentes. Folia petiolo 
1-lJ lin. longo suffulta, elliptico-obocata, |-1^ poll, longa, 4-6 lin. lata, basi breviter 



140 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

in petiolum acutata, margine valde revoluto, basi integra a medio vel in parte 
te?'tid siiperiore minute glanduloso-serrata, apice obtusa, acuta vel i^etusa, scepe 
mucromdo terminata, coriacea, supra glabra, rugulosa, et secundum nervos laterales 
valde impressa, subtus in costa parce puberula, caiterum glabra, glanduloso-punctata, 
nervis utrinque 4-6, angulo 70°-80'' ortis, prominulis vel indistinctis. Stipulae 
minutissimse, subulatae. Elores mascuU in cyniis paucijloris, interdum ad fiores 
solitarios redactis^ 3-4 lin. longis, bracteatis puberulis divspositi ; pedicelli 2 lin. 
longi. Sepala rotundata, ^-f lin. longa, glabra, minute ciliata. Petala alba, basi 
connata, rotnndata, l^~l|lin. longa. Eilamenta antheris multo breviora. Ovarium 
rudimentarium, globosum. Plores feminei solitarii, in axillis foliorum superiorum 
vel in ramulis abbreviatis ; pedicelli puberuli, sursum incrassati, 4 lin. longi. Sepala 
et petala ut in flore masculo. Ovarium ovatum ; stigma sessile, breviter trilobum. 
Bacca nigra, tripyrena, globosa, 3 J lin. dimetiens. 
At 11,000 feet {Haviland, 1087). 
Ilex revohda is very well marked by tlie revolute leaves, the apex of which is usually 

obtuse, with a few minute, sharp teeth, and a more or less distinct mucro. It is 

nearest allied to I. crenata, Thunb., and I. rugosa, Max. 

Ilex vacciniifolia, Stapf, n. sp. Erutex. Eamuli stricti, dense foliati, angulati, 
nigrescentes, juniores minutissime atque parcissime puberuli. Folia petiolo l-lj 
lin. longo suffulta, ohovata vel ohlonga, plerumque 4-7 lin. longa, 3-4 lin. lata, basi 
acuta, apice retusa, integra, coriacea, glabra, supra nigrescentia, nitida, subtus opaea, 
glanduloso-pnnctata. Elores (masculi tantum noti) in racemis axillaribus 4-6 lin. 
longis 10-20-/lo)ns dispositi, apicem versus magis congesti ideoque interdum subum- 
bellati. Pedicelli graciles, 1 lin. longi. Sepala 4, glabra, basi connata, rotunda, 
minute ciliata. Petala basi connata, rotundato-ovata. Pilamenta brevia. Ovarii 
rudimentum stigmate sessili lobulato coronatum. 

a. Var. CAMPTONEURA. Polia o\)OY2it3^i subplana, nervis subbasalibus sub margine apicem 
versus productis distinctis, caeteris plus minusve inconspicuis. 
At 6600 feet, on the top of a ridge {Haviland, 1186). 

/3. Var. SLBENERVis. Polia magis coriacea, plerumque oblonga, marginibus reciirvis, 
enervia vel nervo uno alterove subtus prominulo. 

At 8800 feet {Raviland, 1114). 

Like the preceding species, nearest allied to I. crenata, Thunb., from which it differs 
in the entire margin and the retuse apex as well as in the flowers, which are arranged in 
regular, often subumbellate racemes. The variety subenervis is extremely like some of 
the small-leaved specimens of I. Walkeri, Wight, so far as the habit is concerned ; but 
the marginal nerve, instead of forming a distinctly broken line, is steadily curved, and 
the inflorescence is totally difl'erent. 

CELASTRINEiE. 

MiCROTROPis RAMiFLORA, Wight, Icon. t. 977. 
At 8000 feet {Raviland, 1110). 
Distribution : Ceylon and Western Ghats. 



DE. O. STAFF OX TJIE FLORA OF MOUNT KIXARALU. 14.1 

Perrotettia alpestris, Losener, in Engl. Natiirl. Pllanzenfam. iii. Ih. 5. Abtli 220. 

, Tar. philippixexsis, Stapf.— Syn. P. jjhilippinensis, Losener, I.e.; Canjospermum 

pJiiUjjpinense, Yidal, Ptev. PI. Filip., 89. Flores tetram<>ri. 
A small tree on the Daliombang River, alt. 3000 feet {Ilariland, 1389), 
Distribution (of species) : Malaya, from Sumatra to the Phili])pines. 
I find the flowers in Vidal's type specimen of Caryospermnm p/ulippinense tctramerous, 
as they are in theKinabalu plant; while all flowers which I examined from Sumatra and 
Javan specimens of P. alpestris, Losener, were pentamerous. But apart from tliis I am 
not able to indicate any differential characters which mii^ht be used for separatinj^ these 
two supposed species. Losener puts P. phil'q)pincnsis and P. alpestris into his section 
Caryospermum, but neither of these specimens has solitary ovules. I dissected a great 
many flowers and fruits and found tiie ovary always 2-celled, with 2 erect ovules. In 
some cases, however, a slin;ht ridge may be seen protruding between the seeds like a false 
septum, when they fill up the cells. Yet I rather derived the impression that this was 
the mere mcchaniail consequence of the pressure of the growing seeds against the soft 
ovary-walls. The flowers are polygamous, and show all intermediate states from male 
to female flowers. In the latter the stamens are replaced by almost petaloid, though 
small, staminodes w ithout any trace of anthers. P. alpestris approaches very closely 
P. satidwicensis, A. Gray. 

Salacia latjrifolia, Stapf, n. sp. Frutex glaberrimus. Rami novelli subtetragoni, 
nigricantes, annotini albidi, teretes. Folia pctiolo 2 lin. longo vel breviore sutfulta, 
lanceolata, 2J-3^ poll, longa, 8-10 lin. lata, basi sensim attenuata, apice breviter 
acuminata, margine integerrimo leviter undulato, crasse coriacea, fusco-olivacea, 
nervis lateralibus utrinque 6-7, arcuatis, utrinque prominulis, reticulatione lax^, 
indistincta. Inflorescentiae axillares vel extra-axillares, fasciculata?, O-lO-florae ; 
pedunculi brevissimi; pedicelli 1 lin. longi. Calyx breviter cupularis, crasse et 
breviter 5-dentatus. Petala in alabastro anguste imhricata, oblonga, 1 tin. longa, 
crassa, in dorso carinata et glauca, suberecta. Stamina 3, ovario medio vel supra 
insidentia; filamenta late triangularia, apice demum recurva; antherai subquadratae, 
J lin. longse. Styli 3 breves. Ovarium 3-loculare. 
Young jungle on the Ulu Tawaran, alt. 2000 feet (Uatiland, 1370). 
Probably nearest to S. prinoides, DC, but very distinct by the thick, entire, lanceo- 
late leaves, and the carinate thick petals, which overlap each other only very slightly in 
bud. 

AMPELIDE.E. 
Cissrs HASTATA, Miq. Fl. Ind. Bat., Suppi. 512.— Syn. Tit is sagittifolia, M. Lawson, 
in Hook. fil. FL Brit. Ind. i. 615. 
At Koung, alt. 2000 feet [Uaviland, 1372). 
Distribution : Malaya, from Penang and Sumatra to Borneo. 

Planchon, in DC. Monogr. v. (pars 2) 502, quotes " Vitis hastata^ Miquel, Fl. Ind. 
Bat. Supj)lem. 517," whereas Miquel in the passage cited calls the plant "Cissiis hastafa." 
It must be therefore C. hastata, Miq., instead of Planch. 

SECOXD series. — BOTANY, VOL IV. TJ 



3i2 DE. O. STAPP OIS" THE FLOEA OF MOUNT KINABALU. 

SAPINDACEiE. 

GuioA PLEUROPTERis, Kadlk. Ueber Cupania, &c., in Sitz. Ber. bayer. Akacl. Wiss. 1879, 
611. — Syii. Cupania pleuropteris^ (3. apiculata, Hierii, in Hook. fil. PL Brit Ind. 
i. 677. ^ 

Young jungle on the Ulu Tawaran, alt. 2000 feet {Kaviland, 1361). 

Distribution : Malaya, from the Peninsula to Borneo. 

SABIACE^. 

Sabia PARVirLORA, Wall, in Roxb. Fl. Ind., ii. [1821] 310. 
At 5000 feet (Havikmd, 1218). 
Distribution : Tropical Himalaya, from Sikkim to Kumaon. 

ANACAUDIACE^. 

Rhus borneensis, Stapf, n. sp. Erutex, inflorescentia excepta, glaberrimus. Kami 
cortice badio vel nigricante. Eolia petiolo 2-1 lin. Ion go vel breviore suffulta, 
simplicia, obverse lanceolata vel ohlonga^ ad 5 vel 6 poll, longa, ad 2 poll, lata, in 
petiolum attenuata, apice acuminata, demum subcoriacea, nervis lateralibus utrinque 
10--17, intermediis eirciter 3-3J lin. inter se distantibus, in foliis novellis cum venis 
nigricantibus, in maturis subtus prominentibus. Inflorcscentise a basi ramorvim 
hornotinorum 2-5, bracteis parvis deciduis instructa3, anguste paniculatai, masculcB 
1-2 poll. long(B, femine<B demum 3-lJ poll. longcB ; pedicelli ad 1 J lin. longi, pilis 
minutis adpressis conspersi. Elos masculus : Sepala ovata, IJ lin. longa, mem- 
branaceo-marginata. Petala oblonga, obtusissima, vix 1^ lin. longa, reflexa, 
nigro-venosa. Stamina 5 ; filamenta brevissima ; antherae elliptica), J lin. longae. 
Ovarii rudimentum minutum trilobum. Elos femineus : Sepala atque petala at in 
fiore masculo. Stamina plus minusve redacta, effeta. Ovarium oblique ovoideum ; 
stigma sessile, capitatum. Eructus compressus transverse latior, obliquus, 3 lin. 
longus, 4 lin. fere latas ; mesocarpium multistriatum. 
At 6000 feet {Saviland, 1198). 
Inflorescence and flowers are almost exactly as in Hh. succedanea, except that the 

panicles are shorter and the anthers almost sessile. The leaves are simple, without any 

trace of articulation, and exhibit a type of nervation similar to that in Hh. succedanea ; 

the venules, however, are less distinct, probably in consequence of the greater thickness 

of the leaves. 

LEGUMINOS^. 

Desmoditjm Scalpe, DC. Prodr. ii. 334. 

At 4000 feet {Haviland, 1246). 

Distribution : Tropical Africa and Natal ; Madagascar and Mascarenes ; Ceylon and 
Western Peninsula to Concan ; Manipur ; Malaya. 



DR. O. STAFF OX THE FLOKA OF MOIXT KIXABALU. 143 

Baijhinia excurrexs, Stapf, n. sp. (§ Pha}wra). Eamuli i^labri, nii,'rescontes, lenticelUs 
minutis crebre obsiti. Folia petiolo nigrescente, apice ct basi incrassjito, circiter 
\h poll, longo suffulta, subcordato-ovata, 3-1 poll, loni^a, 2A-3 poll, lata, basi levitcr 
cordata sinu latissimo, apice abrupte atque hrecissime acmninafa, norvo medio hnrifer 
excurrente, supra glaberrima, subtus in nervis venisque pilis aureis adpressis minut« 
sericea, nervis 9-11, intermediis 3 approximatis, extimis 2 sippe inconspicuis, 
reticulatione venularuni tenerrima, minutissimtl, sub lente vero distinctissimA. 
Inflorescentia terminalis, racemosa, brevis, parce rufo-pubcrula, l)racteis valde 
caducis. Pedicelli circa 12-11 lin. loni^ri. Calycis tubus cvlindricus, 3-1 lin. 
longus, 1 lin. latus ; lobi lineari-lanceolati, acuminati, 7-8 lin. lonji^i, IJ-IJ lin. lati, 
in alabastro apice in apiculo colia3rentes, sub antbesi reflexi, ut tubus pilis minutis 
adpressis parce rufo-pubescentes. Petala omnia breviter et abrupte unguiculata, 
ungue 1 lin. longo, 2 exteriora elliptica, 10 lin. longa, 4 lin. lata, 3 interiora lineari- 
oblonga, vix 3 lin. lata, omnia pins minusve acuminata, cxtus et intus, ])r;ecipue 
basin versus, fulvo-pilosula. Stiiminnferfili(/ 3 ; iilanienta circa S lin., anthenc 5-0 
lin. longaj ; stamijiodia 2, multo breviora. Ovarium dense fulvo-pilosum, breviter et 
crasse stipitatum, in stylum vix angustiorera sub antliesi duplo longiorem abiens, 
cum CO 8 lin. longum ; stigma magnum, 2 lin. latum. 
At Koung, alt. 2000 feet (Jravilaud, 1382). 

Closely allied to J?, semibijida, lloxb., and B. femujinea^ IJluine, but distinct In- 
undivided apiculate leaves. From B. integrifolia, lloxb., and B. Kockiami^ Kortb., it 
differs considerably in the structure of the flower. 

Baiihixia Burbidgei, Stapf, n. sp. (§ Fhanera). Arbor scandens, 50-00 ped. alta. 
Kamuli dense ferrugineo-villoso-tomentosi^ demum glabrescentes. Folia petiolo 
2-2J lin. longo., rohusto ferrugineo-tomentoso suffulta, oblonga vel elliptico-oblonga, 
5-7i poll, longa, I5-2J poll, lata, basi obtusa, apice plus minusve acuminata, 
acumine obtuso, interdum leviter emarginato, rarius acuto, saepe mucronulato, firme 
membranacea, sujira glabra, subtus in nervis primo ferrugineo-velutina, demum 
glabrata, trinervia, nervis fere ad apicem ductis, supra impressis, subtus promi- 
nentibus, venis transversalibus utrinque circiter 15, in vetustioribus sicpe supra 
impressis, subtus prominulis, distinctis. llacemi terminales, solitarii vel plures ex 
axillis summis udditijferrugineo-villoso-fomentosi; rhachis 2-3| poll, longa, fere totil 
longitudine florigera, vel basi nuda; pedicelli 1-2 J poll, longi, graciles. Alabastri 
limbus subglobosus. Calyx dense ferrugineo-tomentosus ; tubus anguste cylindricus, 
5-7 lin. longus ; lobi ovati, Z\-^ lin. longi, acuti. Petala 5, paulo inaequalia, 
coccinea vel purpureo-rubra, late obovata, majora 12-14 lin. longa, 9 lin. lata, 
abrupte in unguem 1 lin. longum contra eta, obtusa, crenulata, extus ferrugineo- 
sericea, intus subglabra. Stamina fertilia 3, caetera plus minusve redacta, sterilia ; 
filamenta glabra ; antherge fertiles rotundato-elliptica?, fere 1 lin. longse. Ovarium 
ferrugineo-velutinum, mox glabrescens. Legumen lineari-oblongura, 4-5 poll, 
lono-um, 11-15 lin. latum, stylo persistente mucronatum, (siccum) nigrum. Semina 
5-0, lenticularia, 6-7 lin. dimetientia. 

u 2 



144 DR. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

Climbing upon trees on the Tampassuk River near Koung (Burhidge). 

Distribution : Sarawak (JBeccari, 633, 3016 ; Wallace ) ; Labuan ? (Low). 

Closely allied to B. Kockiana, Korth., but distinct by the very copious, almost villous 
indumentum, the short and stout petioles, and by the large flowers, the colour of 
which is much deeper, and perhaps also by the venation of the leaves. Wallace 
describes the colour as " purple red," Burbidge as " scarlet,'' whilst Korthals designates 
the colour of the flowers of B. Kockiana as a bright yellow, turning ultimately into red. 
The typical forms of B. Kockiana, as represented by Korthals's specimen, and of 
B. Burhldgei as represented by the specimens quoted above, are certainly distinct 
enough to be considered as separate species. But there are two specimens here, also 
collected by Mr. Burbidge in North Borneo — the precise locality is not indicated — which 
have rather large flowers and exactly the venation of J5. Burbidgei^ but their indumentum 
and their petioles are exactly like those of B. Kockiana. Mr. Burbidge describes the 
plant named after him in a MS. note thus : " Climbing tree, attaining a height of 50 to 
100 feet, twining around forest trees and falling from their branches in masses of 
green foliage and scarlet flowers." 

PiTHECOLOBiuM BiGEMiNUM, Mart. Herb. Fl. Bras. 115; Benth. in. Trans. Linn. Soc. xxx. 
575. 
In young jungle near the Uiu Tawaran, alt. 2000 feet [Haviland, !1369). 
Distribution : Ceylon and Western Peninsula ; Malaya to the Philippines. 

BOSACE/E. 

Pygeum oocaPwPLM, Stapf, n. sp. Erutex. Bami novelli ferruginco-tomentosi^ mox 
glabrati, vetustiores fusci, lenticellis minutis conspersi. Eolia petiolo 2-3 lin. 
longo, crassiusculo, breviter tomentello suffulta, ovata vel ovato-clllpttca, 1^-2 poll, 
longa, 1-lJ poll, lata, hasi subcordata, apice obtusa, nmrgine breviter recurvo, 
coriacea^ supra in costa nervisque minute tomentella, cgeterum glabra, impresso- 
j^unctata, sublucida, basi utrinque glandula notata, subtus in costa nervisque 
liirtella, cgeterum glabra, nervis lateralibus utrinque 7-8 subpatulis, supra impressis, 
subtus prominulis, venis transversis, supra quidem cum reticulatione gracili distinctis. 
Bacemi pancijlori, ad 5 lin. longi, stricti, ferrugineo-tomentosi ; bracteae ovatae, 
parvae ; pedicelli brevissimi, crassiusculi. Calyx patellaris, ferrugineo-tomentosus ; 
lobi 10 sublineares, remoti, J lin. longi. Petala nulla (?). Stamina 10 ; filamenta 
e basi triangulari subulata cum antheris minutis ^ lin. longa. Dru2)a ovoidea, 
4-4^ lin. longa, 3--3| lin. lata, sparse fulvo-pilosa. 
At 7800 feet (Haviland, 1118). 

l he habit is more like that of P. Wtghiianumi Blume, than of any other sj)ecies; but 
the hairy and ovoid ovary is very different, and the true affinity lies rather w ith P. capitel- 
latnm. Hook, til., and P. htontanimi, Hook, fil., from Tenasserim and Sikkim respectively. 

Ill BUS GLOMEiiATus, Bluuie, Bijdr. 111. 

At 6000 feet {Hamland, 1194) ; at Koung and Kiau {Burbidge). 

Distribution : All over Malaya, from the Peninsula to the Philippines and to New Guinea. 



DR. 0. STAFF ON THE FLORA OF MOUNT KlNABAl.U. 145 

RuBrs LiNEATUS, Reinw. in Blumo, Bijdr. 1103. 

Pakapaka, alt. 10,500 feet, some 15 feet hiwh, in juiii^lc (ILfvilaud, lOGl). 

Distribution : Malaya, from Sumatra to Borneo ; East Himalaya, from Sikkim to 
Mishmi. 



BuBrs Lowii, Stapf, in Hook. Ic. PI. t. 2:289. Bami iuermes^ teretes, nii^resceutes, pi Us 
patulis tomentelUy insuper setulis ^landuligeris parce obsiti, tandem glabrati. 
Stipulae lanceolatte, integrae vel minute denticulatte et basin versus glandulose 
fimbriatse, parce hirsutae vel glabratic, circiter 3 lin. longj«. Polia dii^itata, 
trifoliolata, petiolo tomentello 7-10 lin. longo sufTulta, foliola breviter petiolulata, 
tenninale m.dL]\\% ellij)ticum vel obovatum, magnitudine valde varians, ad ]| poll, 
longum, ad 10-11 lin. latum, basi plus minusve cuneatum, apice acutum vel 
breviter acuminatum, duplicato-serratura, supra nigricans, nervis sericeo-pilosiSj nt 
venul(B firmsversce tenerrimcB plus minusve impressis, subtus brunneum, nervis 
venulisque sericeo-pilosis ; foliola laterdlia oninino consimilia, sed utrinque obtusiora, 
plus minusve asymmetriea, quartA, vel tertiA: parte minora ; folia hinc indradfoUoluni 
soUtarium, late rotuudato-ovalum, indicisum vel plus minusve trilobum redacla, nunc 
parva, f-1 poll, longa, nunc 2-3-plo majora. Plores soUtarii, axillares. Pedicelli 
nutantes, tomentelli, f-1 poll, longi. Calycis segmenta ovata, 2i-4 lin. longa, 
eaudato-acuminata vel apice inciso-dentata, interiora extus in parte in alabastro 
segmentis exterioribus obtecta dense et minutissime tomentella, ca^terum ut Imxce 
pilis paucis longis conspersa, omnia intus griseo-tomentella. Petala rosea, late 
obovata subunguiculata, 2^-3 lin. longa, apiculata. Stamina uniseriata. Syn- 
carpium siccmn, depresso-globosura, 4-5 lin. dimetiens; receptaculum ovoideum, 
dense sericeo-pilosum ; aclia^nia circiter 20, libera, oblique ovata, If lin. longa, 
glabra, foveolato-reticidata; stylus 2 lin, longus, glaber. 
Prom 9000-13,000 feet (Low; Haviland, 1082). 
A very distinct species, wliieli approaches somewhat R. alpestris^ Bl., from which it 

differs in the absence of prickles, the tomentum, the broader leaflets with impressed 

nerves, and the solitary flowers. The achenes of the only ripe fruit I have seen are 

perfectly free and covered with a very thin, hardly fleshy pericarp. 

RuBUS ALPESTRis, Blume, Bijdr. 1103. 

At 7600 feet {Haviland). 

Distribution : Malaya, in high elevations ; temperate Himalayas, from Gurhwal to 
Sikkim, from 7000 to 10,000 feet. 

I have seen only a branch without flowers or fruits, but this agrees so exactly with 
Horsfield's specimen from Mt. Prahu in Java that I do not hesitate to refer it to the 
same species. The Himalayan form constitutes a well-marked variety, characterized by 
the bristles on the calvx. 



1^6 DE. O. STAFF ON THE FLOKA OF MOUXT KINABALU. 

EuBUS KOS^EFOLITJS, Smith, Ic. ined. iii. t. 60 ; Miq. Fl. Ind. Bat. i. pt. 1, 375. 

On the Kadamaian, alt. 2500 feet {Havikmd, 1195) ; alt. 3000 feet (Burhklge). 

Distribution : All over Malaya to New Guinea and South-east Australia ; Indo-China, 
to the Khasia Hills and the temperate Himalaya ; Japan ; South Africa and Madagascar 
(introduced?). 

M. roscefolius is " cultivated around the upland villages of North Borneo near Kina- 
balu," according to Bur bid ge, as it is in many parts of the Tropics ; whence it is very 
difficult to trace the original area. 

EiBus FEAXiNiroLiiTs, Poir. Diet. vi. 242. 

Near Kiau, alt. 3000 feet {Low ; Haviland^ 1376) ; on the northern slopes to 5000 feet 
(Loic). 

Distribution : Throughout the Malay Archipelago, from Java to the Philippines and to 
Tidor. 

PoTENTiLLA LEUCONOTA, D. Don, Prodr. PI. Ne])al. 230. 

, var. BOENEEXSis, Stapf. Polia utrinque dense argenteo-sericea. Bracteolse, sepala, 

petala, stamina 2-3-plo majora, cseterum simillima. Achaenia paulo majora, magis 
elongata, sub apice sericeo-pilosa. 

In crevices of the rocks, and on dry sheltered places from 11,000 to 13,000 feet {Loio ; 
Ilavilcmd, 1058). 

Distribution (of species) : Alpine Himalaya, east of Gurhwal to South-west China 
(Yunnan, Mt. Omei) ; Borneo ; New Guinea. 

This plant is mentioned by Sir Joseph Hooker in his PL Brit. Ind. ii. pt. 2, 352, 
as found on Kinabalu, but referred to P. leuconota, D. Don, without any further 
remark. There is no doubt with regard to the extremely close affinity between the 
Himalayan and Bornean plants. But from the more complete material which is now in 
my hands it appears to me necessary to distinguish the latter from the former by the 
characters given above. 

Sir P. V. Mueller indicates a variety ^' pcfpuana " of the same species from the Owen 
Stanley Banges and from Mt. Musgrave, from 8000 to 13,000 feet (Trans. Boy. Soc. 
Victoria, vol. i. pt. 2, 5). The specimen of it which I saw agrees perfectly with the 
Kinabalu plant in habit, except the poor inflorescences, whilst the flowers are inter- 
mediate between those from Kinabalu and from the Himalaya, and the achenes are 
glabrous and even more elongate than in the variety borneensis. 

PoTENTiLLA MooNiANA, Wight, Icon. t. 233. 

, var. KiNABALTJENSis, Stapf. Polia basalia foliolis majoribus utrinque 16-20, supra 

glabris, subtus in nervis sericeis, in dentibus pilis longis patulis ciliatis, foliolis inter- 
]:)ositis minutis, interdum suppressis. Bracteolae sepalis angustiores, lequilongoe, 
integrtc vel rarius minute trideutatae. Achsenia paululo majora, nervis prominulis. 
In swampy, sheltered, sunny spots, at 11,000 feet (Ilaviland, 1056). 
Distribution (of species) : Ceylon ; Himalaya, from Kumaon to Sikkim ; Khasia Hills; 
Borneo. 



DR. 0. STAFF OX THE FLORA OF MOUXT KINABALU. 147 

This variety is much nearer to the Ceylon and the Khasia form than to that of the 
Himalaya, which constitutes a well-marked variety if not a species, and corn^sponds 
with Wallich's P. polyphjlla (Cat. n. 1020, not n. 1030). Numher 1030 of Wallich's 
herbarium is his P. harbata {~P. polyphylla, var. harhaia Lehm. llcvis. rotont. 54), and 
is identical with the Ceylon j^lant which was described as P. Moonlana by Wii^'ht. The 
specimen from Kinabalu is chiefly characterized by the numerous larj^er leaflets and 
the very small interposed ones, and by the long spreading hairs on tlieir margin. 

PoTENTiLLA PAKVULA, Stapf, in Hook. Ic. PI. t. 2293 {^PoJyphyllcc^ Lehm.). Caulis 
valde abbreviatus, parte epigoe^ ^-2 poll. longt\, inferne foliorum atque stipulorum 
residuis sericeo-lanatis dense vcstitus, superne arete foliatus, sini[)lox vel bre- 
vissime parceque ramosus, apicem versus ramulos uuifloros edens. Hamuli floriferi 
erecti, 1-2-foliati, 1-1^ poU. lougi, pilis sericeis longis plus minusve patulis vestiti. 
Stipulae petioloadf adnatae, oblongjie vel oblanceolatse, t-0 lin. longae, lJ-2.} lin. latic, 
acutee, scariosa), extus sericeo-pilosa), intus glabrae, in ramulis floriferis minores, plus 
minusve bilobas vel bilidiie. Folia pinnatifida, rhachi ad J-1 poll., vel in speci- 
minibus nanis et in ramulis floriferis tantum ad 1-3 lin. niulj\, patule soriceo-pilosA-, 
^-4 poll, longa, 4-8 liu. lata ; foliola basin versus decrescentia, utrinque 10-12, 
in specimiuibus nanis fere imbricata, sessilia vel breviter petiolulata, roluadato-ovata, 
superiora 2-3 lin. longa, 11-2^^ lin. lata, utrinque S-1-dentata vel in latere basin 
versus spectaute subintegra, denticulis ovatis obtusiusculis (si pauca, semper sub 
ipso apice), rigidiuscula, demum castanea, supra glaberrima, nitida, infra in costula, 
atque in margine pilis longis paucis obsita ; in foliis longioribus sacpe foliola pauca 
minuta Integra interjecta; folia caulis floriferi consimillima, plcrumque valde 
redacta. Involucelli bracteolce rotundato-ellipticce^ ^2~^^ 1^^* longoe, obtusa^, ut 
sepcda ovata, acuta, sequilonga, parce sericeo-pilosae. Petala obovata, 2-2| lin. longa, 
aurea. Stamina 10, bina sub mwquoque j^^talo. Torus dense sericeo-pilosus. Ovaria 
glabra, breviter oblique oblonga, stylo subbasali, superne leviter incrassato. 
In boo"o>v, sheltered, sunny places, alt. 11,000 feet [Low ; Haviland, 1057). 
P. parvula is nearest allied to P. Mooniana, Wight, and Dr. Haviland suggests that it 

niio-ht even be an extreme form of it. 

Stranv^sia integrifolia, stapf, in Hook. Ic. PI. t. 2291. Frutex. Hamuli nigrcscentes, 
novelli plus minusve tomentelli, demum glabrati, cortice plumbeo. Folia petiolo 
4-6 lin. longo, supra canaliculato, rubescente, tomentello vel glabrato suffulta, elliptico- 
oblonga, 2-3 poll, longa, 10-12 lin. lata, basi breviter acuta, apice acuta vel sub- 
acuminata, rarius obtusiuscula, margine integro, supra nitida, in nervis pilosula, 
demimi glabrata, subtus glabra, in margine primo minute lanuginosa, mox glabrata, 
coriacea, nervis lateralibus 7-9, tenuibus, vLv prominuUs, valde obliquis. Corymbi 
tomentelli, 20-30-flori. Pedicelli 1-2 lin. longi. Calyx turbinatus, tomentellus ; 
tubus IJ lin. longus ; lobi persistentes, triangulares, \ Im. longi, acuti. Petala 
rotundata, 2 J lin. dimetientia, alba. Stamina circiter 20. Ovarium ad \ cum 
calvcis tubo connatum, 5-loculare ; stvli 5 ad } connati ; stigmata apice iucrassata, 



148 DE. 0. STAPE ON THE FLORA OF MOUNT KINABALU 

truncata. Pructus (baud perfecte matnrus) 2J lin. dimetiens, vertice liemispliajrico, 
e tubo calycis exserto, leviter 5-sulcato. 
From 11,500 to 13,000 feet {ILaviland, 1071). 

So closely allied to Stranvcesia glaucescens, Lindl., from the Himalaya and the Xhasia 
Hills, tbat it might be almost considered as a variety of it. But the corymbs are stouter, 
denser, and less tomentose, and the leaves absolutely entire, with fewer and fainter nerves, 
which are curved more forward than is the case in ;S'. glaucescens. A specimen of 
S. Davidiana, Hecne., from Szechuen [Henry, 8953), comes also extremely near, but the 
lateral nerves are about twice as numerous and straighter. 

SAXIPEAGACE^. 

BicimoA FEBRiFUGA, Lour. M. Cochinch. 301. 

Koung, alt. 2000 feet {Biirhidge ; Ilaviland, 1375). 

Distribution : Malaya, from the Peninsula and Sumatra to Celebes and the Philippines ; 
Indo-China to Mt. Omei', Hupeh and Pokien ; Burma, Khasia Hills, and Himalaya 
to Nepal, 

Itea macrophtlla, WaU. in Eoxb. PL Ind. ii. [1824] 419. 

In young jungle on the Ulu Tawaran, alt. 2000 feet [Haviland, 1368). 

Distribution : Malaya, from Java to the Philippines; Khasia Hills and Eastern Himalaya. 

PoLYOSMA HooKERi, Stapf, in Hook. Ic. PL t. 2296. Prutex. Hamuli novelli minute pu- 
beruli, mox glabrati, nigricantes, vetusti lenticellis pallidis obsiti. Polia opposita, 
petiolo 3-5 lin. longo, tereti suffulta, ohovata vel clUptica, 1-1 J poll, longa, |-1 poll, 
lata, interdum, multo minora, basi breviter acuta, apice obtusa, emarginata, margine 
anguste revoluta, coriacea, glaherrima, nigricantla, supra lucida, nervis lateralibus 
utrinque 4-6 angulo subrecto ortis. Inflorescentia terminalis, hrevissimc cyynuloso- 
racemosa, pauciflora. Pedioelli 2-2|^ lin. longi, puberuli, sub calyce 2-3-bracteolati. 
Bracteolse lineari-lanceolatae vel subulata?, 1 lin. longae. Calyx turbinatus, IJ lin. 
longus, puberulus, lobis triangularibus, 1 lin. longis, persistentibus. Petala linearia 
vel lineari'lanceolata, 1 'poll, longa, vix \h lin. lata, extus, imprimis basin versus, 
flavido-sericeo-tomentella, intus rosea villosula. Stamina 4 ; filamenta linearia, 6-7 
lin. longa, villosula ; antherse lineares, vix 3 lin. longae. Stylus 9 tin. longus, glaher ; 
stigma capitulatum. Pructus niger, ellipsoideus, 4-5 lin. longus, calyce persistente 
coronatus, unispermus. Semen globoso-ellipsoideum ; embryo minimus in ima basi 
situs. 

Prom 8000 to 10,500 feet, commonest at 10,000 feet [Low ; llarAUnid, 1068). 

A very distinct species, with large flowers. 

PoLYOSMA BRACTEOSUM, Stapf, n. sp. Prutcx altus. Ptamuli juniorcs tomento dcnso 
jlavido vestiti, vetustiores nigricantes. Polia opposita vel subopposita, petiolo 
robusto, tomentello, |-1 poll, longo suffulta, elliptica, 3-4J poll, longa, 2-2^ poll, lata, 
basi rotundata, apice obtusa, Integra vel sub apice utrinque 1-2 denticulata, revoluta, 



DE. O. STAFF 0^' THE FLOEA OF MOUNT KIXARALU. 149 

coriacea, supra primo tomentosa, mox glabrescentia, exsiccando nigricantia, lucida, 
subtus flavido-tomentosa, toinento taudem plus niinusve evauido, ncrvis lateralibus 
utrinque 8-10, utriuque prominentibus, laxc reticulata. Ilacemi stricti, laxiusculi, 
dense tomentosi, ad 5 poll, longi; bmctece Uneares circa 5 lin. longce \ pcdiccUi 
brevissimi ; bracteolae bracteis consiuiiles. Calyx cupularis, IJ lin. altus, tomcn- 
tosus, lobis ovatis. Petala fuscescentia, e basi ovatd late liiteariay erecta, apice tan- 
tum pateutia, 6 Im. longa, supra basin 1 J lin. lata, extus et Intus in parte superiorc 
tomentosa. Eilamenta linearia, basi excepta papillis albis loiii^iusculis vestita, 2-2| 
lin. longa; antheroB liueares, li lin. ionga3. Stylus crassiusculus, cylindricus, 1 lin. 
longus, sericeo-pubescens ; stigma minute bilobum. Fructus oblongo-ovoidcus, 
tomentosus, 5 lin. longus. Semen solitarium, 3 lin. longum. 

At 9000 feet, (Maviland, 1103). 

Allied to P. mutabilcy Blunie, but distinct by its large flowers and bracts and the more 
copious tomentuni. 

DEOSERACEJE. 

Dbosera spathulata, Labill., PI. Nov. Koll. i. 71), t. 106, i". i. 

Maripari Spur, from 5000 to 5500 feet {Low, Ilaviland). 

Distribution: South-east China, Borneo, Pliilippines ; East Australia from Queens- 
land to Tasmania ; New Zealand. 

I am not able to distinguish the Chinese form which has been described as 
J). Loureiri, Hook, et Arn. (Bot. Beech. Voy. 107, t. 131) from the Australian D. spa- 
thulata, and consider it to be quite the same plant, as Bentham has already suggested 
(PL Austral, ii. 459). 

HALORAGINE^. 

Haloragis micrantha, R. Brown in Plinders's Voy. ii. 550. 

In clefts on rocks and open places, in compact tufts, alt. 12,000 feet {JIaviland, 1077). 

Distribution : Prom Australia and New Zealand to Japan, Indo-China, and the Sikkini 
Himalaya. 

MYRTACE-S). 

Leptospermum recurvum. Hook. fil. in Hook. Ic. PL t. 893. 

Prom 7000 to 13,000 feet, commonest above 11,000 feet {Low ; Maviland^ 10G3). 

A shrub, 4-8 feet high below 11,000 feet, but dwarf, a few inches only in height, at 
the top of the mountain, which appears quite white with it when in blossom. 

LEPTOSPERMUii JAVANiCTJii, Blume, Bijdr. 1100. 

A small spreading tree, at 7700 feet {Haviland, 1154). 

Distribution : ^Xlalaya j Moulmein. 

The specimens from Kinabalu exactly agree with the plant collected by Lobb on 
Mt. Thounggyen in Moulmein. They differ slightly from the Sumatran and Javan plant 

SECOND SERIES. — BOTANY, VOL. IV. X 



150 DK. 0. STAFF ON THE FLORA OF MOUNT KINABALU. 

in the less stout habit, the thinner leaves, and the much scantier indumentum of the calyx. 
The difference between Ijeptospernmm javanictim andZ. recttrvum is not greater than that 
between typical L. flavescens, Sm., and some forms referred to var. ohovatum of this species. 
Tlie amount of hairiness in L. recurmim is exceedingly variable, and plants growing close 
to each other may possess either leaves with a dense white silky coat on the back or 
leaves almost without hairs. The same may be said with regard to the calyx. Thus 
the difference is reduced to the stunted habit and the minute leaves. Such minute 
leaves, however, may be found occasionally also in L.jcwanictmi, at the base of the branches 
or on short and poorly developed lateral shoots. This form of leaves evidently has become 
general in L. recurvum, whilst the texture grew firmer at the same time and the margins 
recurved, as often occurs in similar cases with alpine or very exposed forms. 

Decaspermum Vitis-Id^a, Stapf, n. sp. (PL XI. a. figs. 1-9.) Eruticulus. Rami 
graciles, dense foliati, cinereo- vel subnigro-strigillosi, demum glabrati, vetusti cortice 
brunneo. I'olia opposita, petiolo brevissimo suffulta, ohlonga, 3-4 lin, longa, 2 lin. 
lata, basi et apice obtusa^ margine revoluto, coriacea, supra, costula sericea excepta, 
glabra, impresso-punctata, mox glaberrima, lucida, snbtus in costula et in marginibus 
longeargenteo-sericeo-pilosa demum plus minnsve glabrata, minute nigro-punctulata, 
evenia. Plores axillares, pedicellis 2-2^ lin. longis apice bibracteatis sericeo-pilosis 
suffulti. Calycis tubus turbinatus, argent eo-sericeus ; lobi 4, triangulares vel ovali, 
obtusi vel subacuti, 1 lin. longi. Petala 4, alba, pnrpureo-suffusa, obovato-rotundata, 
ad 2 lin. longa, in margine minute ciliolata. Stamina numerosa, sub 3-seriata, in 
alabastro incurva ; fllamenta petala vix sequantia ; anther{3e minutge, subglobosae, 
loculis longitudinaliter dehiscentibus. Ovarium 5-loculare ; stylus fiiiformis, pilo- 
sulus ; stigma minute capitatum ; oviila solitaria, fere e basi erecta. Bacca (imma- 
tura) subglobosa, parva ; semina 5, reniformia. 
Maripari Spur, alt. 5500 feet {Mavlland, 1261). 

A very distinct species, characterized by the small oblong leaves and the solitary ovules. 
Young leaves are sometimes very like the minute leaves which now and then occur in 
Decaspennum paniculatum amongst the normal ones, or like those of D. Blancoi, Vidal, 
to which the species is nearest allied ; but they soon become much firmer, whilst the 
margins curl in, and then they resemble very much those of Vaccinium Vitis-Idcea. 
I found the ovules always solitary, rising from a little above the base of the cell. Never- 
theless, the back of the cell may sometimes be seen intruding slightly into the cell, 
forming a faint central ridge, thus indicating the formation of a false septum, which is 
more or less normal in other species of Decaspermum. 

Decaspee-HUM PANiciLATrM, Kurz, in Journ. As. Soc. Beng. xlvi. (1877) part ii., 61. 

At Koung, alt. 2000 feet {Haviland). 

Distribution: Throughout Malaya, from the Peninsula to the Philippines and New 
Guinea. 



DE. 0. STAFF OX THE FLORA OF MOUXT KJNAHALU. 151 

MYRTrs FLAVIDA, Stapf, in Hook. Ic. PL t. 2290. Arhor parva velfrutex, ramis dimncatis. 
Eamuli novelli plus miiiusve dense albo-villosi, vetustiores cortice niii^riejintc \v\ 
griseo. Folia opposita, patula vel reflexa, petiolo ^-| lin. l()ni,'o sullulta, ovato- 
lanceolata, 9 lin. longa, 2-3 lin. lata, basi rotiindata, obtusa vel subacuniinata, 
acumine obtuso,.marginibus plus minusve revolutis, coriacea, supra nitidula, ])iini() 
villosula, mox glaberrima, subtus densius atque diutius villosa (imprimis in costA) et 
glanduloso-punctulata, nervis lateralil)us 2 subbasalibus sub margine fere ad apicem 
ductis, coeteris subrectis, sujn-a tantuni visibilibus et ut priores levissime impressis. 
riores axillares, solitarii vel 2-3 fasciculati. Pedicelli villosi, i-2 lin. longi, vel 
subnuUi. Calycis tubus turbinatus, 1 lin. longus, villosus; lobi 4, suberecti, post 
anthesin conniventes, vix tubo breviores, triangulares. Stamina numerosa, biseriata ; 
filamenta in alabastro, interiora quidem, intiexa, in flore aperto 2-3 lin. longa -, 
antlierse basi affixse, rotundato-ovatoL% \ lin. longa\ Stylus filiformis ; stigma puncti- 
forme. Ovarium biloculare ; j)laceiitje paulo promiimhc ; ovula in utroque ioculo 
circa 10-14, biseriata. rructiis baccatus, calyec persistente coronatus, globosus, 
lj-2 lin. dimetiens, niger. Semina arete compressa, 10-12 in utroqu(» Ioculo, 
subreniformia, angulata, i)allida, testj'i crassiuscula durii,. Embryo a me ikmi visus, 
sed cavitas seminis curvata, angusta. 
Maripari Spur, alt. 5500 feet; alt. 7700 feet {ILn-ihnid, 1155). 

This is a very remarkable species. It is the first Myrtus found in the ArchijK^lagc. 
The habit is almost that of certain South-American species, as M. myrlcoidesy II. B. K., or 
M. microphylla, Humb. et Bonpl. One specimen with very small and very re volute leaves 
comes more closely to M. rufo-punctatus, Brongn. et Gris, from New Caledonia, whilst 
the nearest ally among the Australian species is perhaps 31. 3Ict)'osidrros, BailL, from 
Mount Bellenden Ker in Queensland, a plant which is known only from barren branches. 

Teistania ellipticA, Stapf, n. sp. (§ Eu-Tristania). (PI. XI. d. figs. 14-17.) Frutex 
humilis. Hamuli nigricantes, crassiusculi, glaberrimi, demum cortice irregulariter 
soluto, pallidiores. Folia alterna vel summa opposita, approximata, petiolo latiusculo 
circa 1 lin. lougo suffulta, elUptica vel obovata, l^-2f poU. longa, 1-lf poll, lata, basi 
petiolum acutata, apice obtusa vel obttisissima, valde coriacea, glaberrima, supra 
lucida, subtus opaca, costtl supra impressa, subtus crassiuscula prominente, nervis 
lateralibus supra prominulis,utrinque 15-18, 2 lin. inter se distantibus, parallelis, rectis, 
sub ipso margine cum nervo marginali junctis, venulis supra interdum prominulis 
laxe reticulantibus. Cyma3 terminales et axillares, breves, congestoi, pedunculo vix \ 
p>oll. longo suffultse, bracteatae, glabrae. Bractea^ ovatae, J-| lin. longae, valde caducaB. 
Pedicelli 1-1 J lin. longi vel vix ulli. Calyx depresso-turbinatus, 1-1 J lin. longus, 
dentibus brevibus triangularibus, extus glaber, intus minute albido-tomentellus. 
Petala obovata, transverse latiora, vix \ lin. longa. Stamina circa 15 ; filamenta terna 
ad medium connata, exteriora cuj usque fasciculi breviora. Ovarium subsuperum, 
3-loculare, globosum, villosulum ; stylus e calycis tubo paulo exsertus ; ovula reflexa, 
numerosa. Capsula subglobosa, imd tantum basi calycis tubo adnata, exserta, ad 2 lin. 
longa. Semina 6-8 in quoque Ioculo, \ lin. longa, ala oblonga fere triplo longiore. 

x2 



152 DR. 0. STAPP ON THE FLOEA OE MOUNT KIN AB ALU. 

Maripari Spur, alt. 5500 feet (Havilmid, 1257). 

Very distinct by the crowded and very coriaceous, mostly broad-elliptic leaves, and the 
robust and short inflorescences. The capsule is of the same size as in Tristania WigJitiana, 
Griff., or T. hurmanica, Griff. The nearest allied species which I have seen is represented 
by a specimen (still undescribed) collected by Beccari in Saraw^ak (no. 2224). 

Tristania bilocularis, Stapf, n. sp. Prutex. Kamuli novelli glaberrimi, nigricantes, 
vetustiores cortice pallido vestiti. Tolia alterna vel summa subopposita, petiolo 
3-5 lin. longo suffulta, obovato-oblonga, lJ-2 poll, longa, |-1 poll, lata, cuneatim in 
basin attenuata, obtusa, marginibus revoUitis, coriacea, glaberrima, supra subnitida, 
subtus opaca, nervis secundariis utrinque circa 12-15, IJ lin. inter se distantibus, in- 
distinctis, marginalibus sub ipso margine vix visibilibus. Cjmsepaucijlone, terminales 
vel ax.iUsiYes, pedunciclo 2-3 lin. longo. Pedicelli 1 lin. longi, glabri. Calyx depresso- 
turbinatus, 1 lin. longus, deutibus minutis triangularibus, intus minutissime tomen- 
tellus. Capsula elongato-globosa valde exserta, ima tantum basi calycis tuboadoata, 
vix Ij lin. longa, septicide bivalvis. Semina 2 in utroque loculo, pendula, IJ lin. 
longa, elliptica, superne alata. 
At 6600 feet [Haviland, 1199). 
I have not seen the flowers of this species; yet the scars left by the stamens between 

the calyx-teeth show' even in the mature state the three vascular bundles of each staminal 

bundle. The capsules are smaller than in T. Wighticma, Griff., to which it is' perhaps 

more nearly allied than to any other species. 

Eugenia kinabaluensis, Stapf, n. sp. (§ Syzygimn). (PI. XL b. figs. 10-12.) Prutex. 
Hamuli novelli tetragoni, vetustiores teretes, cortice brunneo in lacinias angustas 
longas soluto. Polia superne valde appro ximata, petiolo vix 1 lin. longo suffulta, 
rotund at o-ovata vel orbicularia, 4-6 lin. longa, 3-6 lin. lata, basi rotundata, apice 
valde obtusa vel subemarginata, marginibus revolutis, coriacea, supra opaca, subtus 
sublucida, pallidiora, glaberrima, costa supra impress^, tenui, nervis lateralihiis tenuis- 
simis 6-8, supra tantum plus minusve distinctis. Cymae terminales, pauciflorse, 
breves, vix e foliis exsertse. Calycis (juvenihs) tubus breviter suhcyli^idricus, IJ lin. 
longus ; lobi 4 ovati, IJ lin. longi. Petala libera, elliptica vel rotunda, 1 lin. longa. 
Ovarium 3-loculare ; ovula numerosa in quoque loculo. 
At 8800 feet (ITaviland, 1112). 

This, like the following species of Eugenia, is not represented in a quite complete state 
in Dr. Haviland's collection. I should therefore have refrained from describing them 
as new species if they had not been of so characteristic a type that I could not have 
overlooked them in going through the material in the Kew Herbarium and through the 
descriptions of those comparatively few species of tropical Asia which are not represented 
in oar collections. E. kinabalucnsis has quite the habit of E. rotimdifolia, Wight, from 
which it may be distinguished by the very indistinct and looser venation of the leaves, and 
the longer, more cyhndrical calyx. The difference in the venation, however, is only one 
of degree, as the lateral nerves in E. kinabaluensis are hardly less numerous ; but they 
are alternatively so faint that thev do not count. 



DR. O. STAFF OX THE FLORA OF ,MOUXT KIXABALU. 153 

Eugenia ampullaria, Stapf, n. sp. (§ Syzygiiim). (PI. XI. c. tig. 13.) Frutex. l{;iimili 
subteretes, 'brunnei, rami crassi, albescentes. Folia vahle approximata, petiolo cras- 
siusQulo nigresceiite 1 liu. longo suff'ulta, rotunchdo-ovata, ad 1-lJ poll, louga, 10-1 1 
lin. lata, basi late rotundata, apice subacuta vel obtusa, in margine paulo rccurva, 
coriacea, supra lucida nigrescentia, mhtus opaca pallide olimcea^ glaborrinia, costiH 
supra impressa, nervis lateralibus iitrinque 15-20, angulo fere recto ortis, rectis, 
idrinqiie promimdis, imprimis subtus conspicuis, nervo raarginali sub ipso maro-ine 
ad apicem ducto, venulis plus minusve conspicuis, laxe anastomosantil)us. Cyince 
terminales, pauciflorse, hrevissima^, vix e folds exserUr, rohmtce. Fructus globosnSy 
5-6 lin. dimetiens, cahjcis tuba persistente 3 lin. I on go basi constricto coronatm. 
Semen solitarium, globosum, 3 lin. dimetiens ; cotyledones crassissimi, liemispliEerici, 
subinagquales. 
At 11,500 feet {Haviland, 1096). 

E. ampidlaria is closely allied to a species collected by Tlnvaites on Adam's Peak and 
erroneously referred by bim and Dutbie to B. calophyUifol'w, Wigbt, a plant represented 
in the Kew Herbarium from the Western Peninsula only. This is distinguished cliii^lly by 
the larger fruits, which are (immature) 5-0 lin. long and bottle-shaped, with but a sliglit 
constriction at the neck. The part below the constriction is obovoid, tapering into the 
base, that above cylindric and suddenly widened into the margin, which is minutely 
toothed or almost entire. All the fruits of the Ceylonesc specimen are immature, but in one 
or two which are more advanced the lower part is almost globose, thus approaching the 
shape of the fruit of B. ampidlaria. The inflorescence is also of the same character as in 
the Ceylonese plant mentioned, which is now ^. Fergusoni, Trim., but, like all other parts, 
much stouter. The leaves of E. Fergusoni are also very similar, but smaller and not 
glossy, and their veins are not prominent above. 

EuGEi^iA Mtrtillus, Stapf, n. sp. (§ Syzygium). Frutex. Ramuli quadranguli, rami 
cinerei, graciles, plus minusve teretes. Folia approximata, petiolo 2-2\ lin. longo 
suffulta, oblongo-lanceolata, |-1^ jooU. longa, 4-6 lin. lata, cuneatim in petiolum 
attenuata, obtuse acuminata vel obtusa, margine tenuiter recurvo, glaberrima, coriacea, 
utrinque opaca, subtus pallidiora, nervis lateralibus primariis utrinque 10-12, 1-3 
tenuioribus ante nervum marginalem evanidis interjectis, omnibus plus minusve 
parallelis, utrinque prominulis, subtus magis distinctis. Cymae terminales, breves, 
graciles, 10-20-floraB, pedunculo 3-5 lin. Jongo suffulta?, minute bracteatae. Calyx 
(juvenilis) obconicus, suhangustus, lJ-2 lin. longus, lobis 1 brevibus, ovatis. 
At 8800 feet [Haviland, 1109). 
Nearest allied to E. cnneata, Wall., and E. rubicunda, Wight. The leaves come 

pretty near to those of E. rubicunda in venation and form, but they are smaller aud 

comparatively broader. The inflorescence and the young flowers, on the other hand, 

agree rather with those of E. cuneata, Wall. 

MELASTOMACE^. 

jVlELASTOMA DECEMFIDUM, Hoxb., Cat. Hort. Bcug. 90. 
Maripari Spur, from 5000 to 5500 feet (Haviland, 1267). 



154 DK. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

Distribution : Throughout Malaya ; on the continent to Moulmein and to South-east 
China. 

Melastoma Beccarianum, Cogn. in DC. Monogr. vii. 356. 

At Kiau, on the Tampassuk River, alt. 3500 feet {Saviland, 1277). 

Distribution : Xorth Eorneo. 

I have not seen the type, but Dr. Haviland's sjoecimens, as well as some collected by 
Lobb near Bruni, Sarawak, belong evidently to this species. The shape of the leaves is 
very uniform, but they are sometimes almost smooth, and sometimes very rough from 
short and stiff bristles. The indumentum of the calyx consists of numerous thick 
processes with long fulvous hairs, which are more or less interwoven and form thus a 
thick coat. 

Blastus Cogniauxii, Stapf, in Hook. Ic. PI. t. 2311. — Syn. OchtJiocharis parviflora, 
Cogn. in DC Monogr. vii. 481. 

Penokok, from 4000 to 6000 feet {Raviland, 1280). 

Distribution : iXortli Borneo. 

The same plant was collected by Dr. Haviland on Mount Matang, at Selabut, and at 
Sepudang, Sarawak. It is a true Blastus. 

Anerincleistus coe-Datus, Stapf, in Hook. Ic. PL t. 2310. Eruticulus (?). Caulis superne 
dense fulvo-tomentellus. Polia petiolo li-3J poll, longo f ulvo-tomentello suifulta, 
late 8iibcorduto-o'oata, ad 9 poll, longa, ad 6 poll, lata, basi leviter cordata, sinu 
angustissimo, apice breviter caudato-acuniinata, Integra, membranacea, supra in nervis 
fulvo-tomentella, prseterea setulis minutis f ulvis aspersa, nigrescentia, infra pallida, in 
nervis dense, in venulis laxius tomentella, 7-9-nervia, venis transversis 2-3 lin. inter 
se distantibus rectis vel subtiexuosis, in margine tenuiter ciliatula. Panicula termin- 
alis pedunculo 4 poll, longo suffalta, 6 poll, longa, tota fulvo-tomentella. Plores 4-1 
ad ramulorum secundi ordinis apices umbellatim congesti, pedunculo ad 2J lin. longo 
vel brevissimo suffulti. Pedicelli circiter 1 lin. longi. Calyx turbinatus IJ lin. longus, 
4-dentatus, dentibus brevissimis, latis, cbtusis, persistentibus, fuivo-tomentellus. 
Petala alba^, rhomboidea, acuta, 1 lin. longa. Stamina 8, sitbcequalia; antherse 
majores sublineares, incur va? vel subrectae, 1^ lin. longse, basi longiuscule biloba?, 
obtusse, utrinque inappendiculafce, minores lineares, vix 1 lin. longse, brevius bilobae. 
Ovarium lamellis angustis ad f cum calycis tubo connectum, vertice coronula quadri- 
loba, Jobis emarginatis, ornatum ; stigma punctiforme. Capsula turbinata, If lin. 
longa, quadrivalvata ; valvse triangulares, emarginatse. 
Penokok, alt. 3500 feet {Haviland, 1281). 
Distribution : North Borneo. 

This plant was collected by Dr. Haviland in Sarawak. It is closely allied to A. Beccarii, 
Cogn., but it differs in its broader leaves and somewhat larger flowers, with a more slender 
calyx. The anthers are in neither species spurred at the back, as is the case in A. hirsuta^ 
Korth., A. Jlelferi, Hook, lil., and in A. Gtnffithii, Hook. fil. In A. Beccarii also the 
anthers are not quite equal, and are partly purple, partly yellow ; but I doubt whether the 



DR. O. STAFF ON THE FLOKA OF MOUNT KINABALU. 165 

latter character is constant. In a s])ecimcn of Aner/'ncleisfKs Bcccaril from 'Mi. IVfatauj^, 
collected by Dr. Haviland, all the ramifications of the panicle are supported and enveloped 
in a young state by densely tomentose ovoid or ovoid-lanceolate bracts, wliich very soon 
come off. In the specimen of ^. Havilandii, the scars of bracts are still to l)c seen on 
the panicle, and it seems to me very likely, from their shape and size, that the bracts 
are also of a similar kind. 

Driessenia glanduligera, Stapf, in Hook. Ic. PI. t. 2201. Herba 4 ped. alta. Caulis 
obtuse quadrangularis, setulisplerumque glanduUgei'is patiilis obsitus. l^oWvi pet iolo 
2-3J j9o//. longo indumento caulis consimili induto suffulta, cordato-ovata^ 4-6J 
poll, longa, 2-3 poll, lata, basi sinu angusto cordata, apice breviter acuminata, 
margine ciliata, membranacea, supra setulis hvcrAbus adprcssis panels aspersa, ecefertnn 
glabra, subtus in riervis setis spcpe glanduligeris patulis et praeterea pilis minutis- 
simis plus minusve adpressis glanduligeris vestita, l-^-nervia, venis transversalibus 
angulo fere recto ortis, anfractis, distinctis, tenuiter reticulata. Paniculae axillares 
a basi ramulosae, hracteatce, circiter 1 poll, longtc, dense glandulose setulosa, flori- 
bus ad ramulorum apices plerumque 3. Bractese ovatce vel lanerolafm vel lineares, 
plerumqe parva?. Pedicelli 1 lin. longi. Calyx breviter tm-binatus, roseus, Ij-l^ 
lin. longus, dentihus 4 snhuJatls, setis patulis glanduloso-capitatis laxe vestitus. 
Petala alba, late ovata, 2J lin. longa, acuta. Stamina 8, aequalia ; anthcnc aurcai, 
lineares, obtusse, basi antice appendicibus binis filiformibus flexuosis, postice calcare 
simili sed breviore auctis. Ovarium ad medium cum calyce connatum, in vertice 
coronula breviter 4-loba, pyramidata ornatum ; stylus flexuosus ; stigma puncti forme. 

At 5000 feet (Raviland, 1174). 

See the remarks under the following species. 

Deiessenia miceothrix, Stapf, in Hook. Ic. PI. t. 2292. Caulis supeme obtuse quadran- 
gularis, inferne teres, ut tota planta pilis minntissimis apice plerumque glanduUgerls 
obsitus. Folia petiolo 3-4 lin. longo sufFulta, valde asymmetrica, ovato-oblonga, 8-9 
poll, longa, 2i-3 poll, lata, basi subrotundata, longe in acumen obliquum attenuata, 
integerrima, membranacea, fere concoloria, tdrinque sub lente valido minutissinie 
pilosula, b-nervia, venis transversalibus angulo fere recto ortis, subrectis, 3-5 lin. 
inter se distantibus. Inflorescentia racemoso-cymosa, ab im^ basi ramosa ideoque plus 
minusve fasciculata, axillaris, 3-4 lin. longa, pedicellis gracilibus, 1-1?. lin. longis. 
Calyx cupularis, J-| lin. longus, dentibus minutis, inconspicuis, latissimis, obtusis, 
persistentibus. Petala alba, late rotundato -ovata, obtusa, i-^ lin. longa. Stamina 
8, subsequalia ; antlierge lineares, ^-f lin. longae, obtusae, basi antice appendiculis 
binis J lin. longis subfiliformibus auctse, postice breviter calcaratae. Ovarium ad J 
calyci adnatum, vertice coronula pyramidata, breviter quadriloba, persistente corona- 
turn ; stylus subflexuosus ; stigma punctiforme. Capsula e calyce persistente ex- 
serta, 1-lJ lin. longa, 4-valvis. 
Penokok, alt. 3500 feet (Haviland, 1279). 
The two species of Driessenia described here have the anthers almost or quite equal ; 



156 DR. O. STAFF ON THE FLOEA OF MOUNT KINABALU, 

they are linear and obtuse, not rostrate, as in Driessenia axantha, Kortli. Otherwise they 
agree in all essential points with the generic characters of Driessenia. The very minute 
glandular hairs, which are dispersed over both surfaces of the leaves in D. microtlirix^ 
may be also found in D. glcmduligera and D. axantha, but in a very limited number and 
only along the nerves. They consist of a stalk, formed by 3-5 cylindric cells, and of an 
ovoid or globose, slightly enlarged terminal cell, which secretes a viscid matter. In 
D. glanduUgera, all parts of the plant are provided with small gland-bearing bristles 
which are homologous to those on the coronula of the ovary of D. axantha, though these 
are rather reduced and very few in number. 

SoNERiLA TENUiFOLiA, Blumc, in Flora, 1831, p. 491. 

Penokok, alt. 3500 feet [Saviland, 1283) ; at 6000 feet (Loio). 
Distribution : Malaya, from Sumatra and Malacca to Eorneo. 

SoNEMLA cuASSiuscuLA, Stapf, u. sp. Hcrba monocarpica, basi lignescens, 7-9 poll, alta, 
interdum a basi ramosa. Caulis setulis crassiusculis adpressis plus minusve dense 
vestitus. Eolia inaqualia, majora pedunculo 1^-2 lin. longo suffulta, ovato-ohlonga, 
hasi acuta vel subrotundata, apice acuta, argute serrata, crassmscula, supra viridia, 
costula setulis adpressis obsita excepta glaberrima, subtus purpurascentia vel intense 
purpurea, in nervis et in parenchymate setulis brevibus crassiusculis arete adpressis 
vestita, S-nervia, nervis subtus tantum conspicuis ; minora multo brevius pcdicellata, 
ovata vel stibrotunda, basi interdum subcordata, 2-3 lin. longa. Cymulse 2- vel l-florae, 
plerumque ad ramorum apices, pedunculo gracili glabro 3-6 lin, longo suffultse; 
pedicelli 2^-4i lin. longi, sub anthesi graciles, demum incrassati, glaberrimi. Calyx 
breviter tubuloso-campanulatus, 2 lin. longus, glaberrimus, dcntibus late obtuseque 
triangularibus, ^ lin. longis. Petala rosea (exsiccata alba), late obovato-elliptica, 
4^-5 lin. longa. Stamina 3 ; antherae e basi bilobae, ovata?, acuminatse, 2 J-3 lin. 
longse. Capsula turbinata, 3 J lin. longa, ut in JS. teuuifolia. 
At SOOO feet {Raviland, 1173) ; Bnrbidge. 

Kearest allied to S. tenuifolia, Blume, with which it has much in common, particularly 
in the shape and structure of the ilowxr and of the capsule. The thick, almost fleshy 
leaves exhibit the same sharp serrature, and where the base is rounded or, in the smaller 
leaves, subcordate, a 4th or 5th basal nerve may be seen, although these nerves are 
always very faint. The leaves of each pair are — except on the uppermost internodes — very 
asymmetric, but a tendency towards this may be observed also in S. temiifoUa, 

SoNEiiiLA KiXABALUENSis, Stapf, n. sp. Hcrba monocarpica, subsimplex, 7-8 poll. alta. 
Caulis herbaceus, purpurascens, quadra ngularis, minute pubescens et praeterea 
parce pilisminutis plerumque glanduligeris aspersus. Polia sequalia vel subuequalia, 
petiolo purpurascente, |-1 poll, longo suffulta, ovato-oblonga, lJ-3 poll, longa, 
|-1 poll, lata, basi lute rotundata, interdum cordata sinu brevissimo angusto, apice 
sensim acuminata, minute crenulata, tenuia, utrinque minutissime verruculosa, 
supra glabra, subtus in nervis venulisque necnon in parenchymate pilis minutissimis 



DR. 0. STAFF ON THE FLORA OF :\IOUNT KIXABALU. 167 

papilliformibas aspersa, Iplo-nervia, duobus nermrum paribus iafimis sub ipsa bast 
oppositis, nervis tertii paris cdtionhus ct plerumque alterniSy venis angulo 50° ortis 
valde prorsus ductis, nervis venisque purpurascentibus. Cymula? cireiiioidoje, sub 
anthesi subumbellatim contractor, terrainales, interduin in ramulo valde abbrcviato 
ideoque specie axillares ; pedunculus gracilis 1-2 poll, longus, piirpurascens, iiulu- 
mento caulis consimili sed parciore vestitus; pedicelli 1-2 lin. longi, superne 
incrassati. Calyx obconicus, 2-2| lin. longus, dentibus late triangularibus, l)revibus, 
minutissime puberulus et parce glandiiloso-setulosus. Potala rosea, oblonga, 4 lin. 
longa, acuta. Stamina 3 ; anthera? e basi ovatii biloba acuniinato-rostrata?, 3-3 1 lin. 
longse. 
In young jungle, at 3200 feet {Haviland, 1282). 

Nearest allied to S. picta^ Griff., from ^vliich it chiefly differs in the shape and the 
nervation of tlie leaves. On one of the leaves there are a few small white patches along 
the midrib, indicating the tendency to becoming variegated, which is tlie rule in S.picta, 
Griff. 

SoNERiLA pi'LcnELLA, Stapf, n. sp. Ilerba monocarpica, subsimplex, ascendens, 3-4 
poll. alta. Caulis subteres, dense minutequc furfuraceo-puberulus, prtcterea setulis 
glanduligeris parce obsitus, rufescens. Polia subjcqualia, petiolo J-f poll, longo 
suffulta, ovata, 1-2 poll, longa, f-1 poll, lata, baM rotundata vel acuta, apice obtusa, 
minute setuloso-serrulata, tenuia, utrinque minutissime verruculosa, supra viridia, 
secundum nervos albo-variegata, costulA, parce furfuraceo-puberuhi excepta glabra, 
subtus in nervis venulisque necnon in parenchymate furfuraceo-puberula, 5-7plo- 
nervia, venis plerumque inconspicuis. Cymulae circinoidese, sub anthesi umbellatim 
congestse, terminates ; pedunculus ^ poll, longus ; pedicelli 2-3 lin. longi, demum 
incrassati. Calyx anguste obconicus, 2h lin. longus, indumento caulis consimili 
vestitus, dentibus brevibus, late triangularibus, acutis. Petala rosea, obovato- 
oblonga, 3^ lin. longa, acuta. Stamina 3 ; antherse e basi ovata biloba rostrato- 
acuminatae (?). Capsula obpyramidata, trigona, 3 lin. longa, laevis, puberula vel 
glabrata. 
At 6000 feet {Haviland, 1200) 

This species is still more closely allied to S. picta, Griff., from which it differs solely 
in the shape of the leaves. I have seen only a single anther, and this was broken in the 
middle. The basal portion, however, was exactly of the same shape as in S. picta. 

SoNERiLA MACULATA, Eoxb. Fl. Ind. i. [1820] 180. 

, var. GLABRATA, Stapf, var. n. Polia subglabra, minute denticulata, non ciliata. 

On the Tawaran River, alt. 2500 feet {Haviland, 128i). 

Distribution (of species) : Prom Nepal and the Khasia Hills to Upper Assam; Borneo. 

Dr. Haviland's specimens, which are mere dwarfs, agree perfectly with S. maculata, 
Roxb., except in the characters mentioned above. 

As the numerous specimens of Roxburgh's S. maculata which I have seen show a 
great uniformity in the ciliate denticulation of their leaves, I thought it expedient to 
distinguish the Bornean form as a variety, though on otherwise trifling characters. 

SECOND SERIES. — BOTANY, VOL. lY. Y 



158 DE. 0, STAFF ON THE FLORA OF MOUNT KIXABALU. 

SoKERiLA BOKNEENSis, Co^. in DC. Monogr. vii. 509. 

, var. SETULOSA, Stapf, n. var. Polia breviora, latiora, tenuiora, supra setulosa, 

subtus in nervis adpresse hirsuta. 

At 3000 feet (Hamland, 1285). 

I have not seen the type of this species, but 1 believe that I recognize it in several 
specimens collected by Dr. Haviland near Kuching. 

The plant from Kinabalu differs in the characters indicated above; yet the speci- 
mens from Kuching show a considerable variability in the shape of their leaves, which 
brings them in one specimen very near to the variety setulosa. The indumentum is also 
exactly of the same kind, but it is much scantier and there are hardly any bristles on the 
upper surface of the leaves. The species approaches closely S. Beccariana, Cogn. 

Phyllagathis "CNirLORA, Stapf, in Hook. Ic. PI. t. 2280. Herba monocarpica, lignescens, 
e basi ramosa, 6-9 poll. alta. Eamuli novelli dense rufo-hirsuti, demum glabres- 
centes, pallidi, subteretes. Folia valde inceqimlia, majoj^a petiolo 3-4 lin. longo 
dense rufo-hirsuto suffulta, oblongo-lanceolata, 1-lJ poll, longa, J poll, lata, basi 
acuta vel rotundata, apice acutiuscula vel subobtusa, a medio in margine minute 
serrata, firmula, supra obscure viridia, secundum costam utrinque 3-5 setidls adpressis 
uniseriatis obsita, preeterea pilis minutissimis, primo dense, deinde laxissime vestita, 
infra pallida, in nervis adpresse rufo-hirsuta, in venis venulisque et partim in paren- 
chymate pilis minutissimis rufis conspersa, 5plo-nervia, pare nervorum infinio ssepe 
tenuiet mAi^i'm.Q.to-, folia minora majoribus consimillima, sed multo minora vel minima, 
brevissime petiolata. Elores solitarii in axillis foUorum, raro terminales ; pedicelli 
crassiusculi, 1^-1 lin. longi, basi bibracteati. Calycis tubus breviter subcampanu- 
latus, subtetragonus, 2 lin. longus, parce pilis minutissimis rufis conspersus ; llmhus 
4i-lohus deciduus ; lohi e basi triangidarl lineares subulati, falcato-recurvi, 1^ lin. 
longi. Petala rosea, obovata, acuta, 4 lin. longa. Stamina 8, aequalia ; tilamenta 
1^ lin. longa ; antherse lineares, 1-1^ lin. longse, obtusse, basi antice minute bilobse, 
inappendiculatse, connectivo postice in calcar breve abeunte. Ovarium subglobosum, 
ad ^ calyci adnatum, in vertice coronula obpyramidata, tetragona ornatum ; stylus 
3 lin. longus ; stigma punctiforme. Capsula subhemisphserica, obtuse tetragona, 
2-2J lin. longa, glaberrima, albida, valvulis breviter bilobis. Seraina oblique ovato- 
oblonga, J lin. longa, granulata, nitidula ; hilum basilare ; raphe tenuis lateralis. 
At 6000 feet (Haviland, 1172). 
See the remarks under the following species. 

Phyllagathis elliptica, Stapf, in Hook. Ic. PI. t. 2279. Herba monocarpica, erecta 
vel e basi repente radicante ascendens, 3-5 poll. alta. Caulis quadrangularis, supernc 
dense hirsutiusculus, et imprimis ad nodos setis longis, pallidis, flexuosis vestitus. 
Polia wqualia vel suba^qualia, petiolo J- J poll, longo, indumento caulis consimil- 
limo vestito suffulta, elliptica, obovata vel oblonga, 2-3J poll, longa, 11 poll, lata, 
basi rotundata, obtusa, minute denticulata, supra primo villosa, demum glabrata, 
subtus m nervis venulisque adpresse hirsuta, 5-7-nervia. Cymuli 2-3-Jiori, sub- 



DK. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 159 

sessiles, terminales vel axillares ; pedunculus crassinscidus, brecissimm, bibracteatus. 
Bmcteae ovatae, 2-3 lin. longae. Pedicolli graciles, 8-9 lin. lon^i, i^laberrimi, demiiiii 
incrassati. Calyx breviter campanulatus, 1| lin. longus, glabor, minute vorru- 
culosus ; limhus decidims, 4-lobus ; lobi ovafo-lanceolati, falcato-vccurti, vix 1 lin. 
longi, apicem versus setulis paucis ciliati. Petala rose^, oblonga vel obovafci, 
5 lin. lon^^a. Stamina 8, sequalia; filamenta 1^ lin. longa; antherse aure^, lineari- 
laneeolatae, IJ lin. longap, obtusiusculoe, basi vix bilobae, antice appendiculat^, 
postice breviter calcaratae. Ovarium ut in P. iniifom. Capsula subJieraisplieerica, 
subtetragona, costulis prominulis 8 ; valvae 4, emarginatae. 
In damp places, from 4000 to 5000 feet {Low ; Eaclhind, 1286). 
A very well-marked species, nearer allied to P. tonkinensis, Stapf, than to the Bornean 
species hitherto known, but differing remarkably in the shape and the indumentum of 
the leaves, the almost sessile few-flowered cymes, the smaller anthers, and the distinctly 
emarginate valves. The extremely short peduncle and the reduction of the involucre 
to two small bracts give the plant a very different appearance when compared with 
P. tonhinemls, but the floral structure is the same, apart from the smaller size and the 
more linear shape of the anthers; and the ripe capsule is almost exactly of the same 
kind in both species. The general appearance, however, of P. umfiora is so very dilFerent 
from that of any I'hyllagathis we know that I should not have hesitated to make it the 
type of a new genus if the structure of the flower and the capsule were not exactly the 
same as in P. elUptica ; besides, the indumentum and the texture of the leaf point likewise 
to Fhyllcigathis. The inflorescence, too, though totally diff'erent at the first glance from 
that of P. rotundifolla, can easily be understood when compared with that of P. elliptica. 
The reduction of the peduncle and of the bracts is carried on still a little further and 
the cyme reduced to a single flower. It is true the habit is very different ; but if we 
remember the great variety which the genus Sonerila exhibits in its vegetative archi- 
tecture, while yet so uniform in its plan, we cannot separate the plant in question from 
Fhyllagatliis only on account of the habit. 

Marumia pachygyna, Korth. Verb. Nat. Gesch. Bot. 242, t. 59. 

On the edge of a clearing, Penokok, alt. 3000 feet {Haviland, 1345). 

Distribution : Borneo. 

The leaves of the Kinabalu specimen agree exactly with those of M. stelMata, Blume, 
which comes extremely near and only diff'ers in the ciHate calyx and the smaller flowers. 

Dr. Haviland says in a note, " petals white." Korthals figures them pink. The 
colour evidently varies from pink to white. 

DissochtETA axnulata, Hook, fil, ex Triana, in Trans. Linn. Soc. xxviii. 83. 
At Bonsol, on the Tawaran Kiver, alt. 1200 feet (Raviland, 1385). 
Distribution : West Malaya, from Penang to Borneo ; Ceram. 

DissocH^TA HiRSUTA, Hook. fil, ex Triana, in Trans. Linn. Soc. xxviii. 83. 

Climbing in trees in young jungle on the Ulu Tawaran, alt. 2000 feet, and near Kiau, 
alt. 3500 feet {Baviland, 1287, 1288). 

t2 



160 DE. 0. STAFF ON THE FLORA OF MOUNT KIN AB ALU. 

Distribution : Borneo. 

Thq ripe fruit is purple and 5-5J lin. long. ' 

Medinilla stephanostegia, Stapf, n. sp. Frutex vel suffrutex. Eamuli teretes, 
novelli setulis 7nifis ftirfmmceo-puberulis brevissimis cestiti, mox glabrati, albidi, 
ad nodos stuposo-barbati. Eolia opposita, petiole 1^ lin. lougo suffulta, ovato- 
elliptica vel oblongo-elliptica, 2-2J poll, longa, 1-lJ poll, lata, basi rotundata, 
breviter obtuseque acuminata, primo utrinque parce setulis ssepe minutis furfuraceo- 
puberulis aspersa, mox glaberrima, 3-5-nervia vel 3-5plo-neivia, sa^pissirae pare 
inlimo nervorum tenuissimo. Panicula terminalis multifloi^a, 3-4 poll, longa^ 
minute parceque furfuracea, t^amis 3-4 verticillatis patentibus, hracteis siiffultls, 
cymas 3-1 -floras bracteatas gerentibus. Bracteae oblongce vel oblong o-lanceolatce, 
3-6 lin. longce, membranacece, albce vel pallide rosecB. Pedicelli 1-3 lin. longi. 
Calyx breviter campanulatus, 2 lin. longas, glaber, limbo integro, membranaceo. 
Petala rosea, late ovata, 4 lin. longa, obtusissima, interdum apiculo minuto terminata. 
Stamina 8, sequalia vel subsequalia ; antlierse e basi biloba lanceolatse 2 lin. longse, 
lobis processubus comiectivi adnatis et brevissime ultra eos productis carinatis, postice 
in calcar deflexura breve abeuntes. Ovarium ad medium calycis tubum adnatum, 
inferne lamellis 4 ad verticem annulo crasso coronatum ductis cum tubo conjunctum. 
Stylus 3-4 lin. longus. Bacca calycis limbo persistente coronata, globosa, 2^ lin. 
dimetiens. 
At 7600 feet {Haviland, 1171). 

Allied to M. himalayana, Hook, fil., and probably to M. eximia, Blume, which I know 
from description only. From M. Mmalayana it differs, apart from the indumentum, 
in the rich, whorled, and multibracteate panicles. It is also related to some of 
the Philippine species, as M. bracteata, Blume, and M. luzonensis, Hook, fil., but less 
clearly. 

Medinilla uuophylla, Stapf, n. sp. Prutex alte scandens. Rami ramulique teretes, 
glabri, demum albidi. Polia ternata, petiolo | poll, longo suffulta, obovato-oblonga, 
2^-3 poll, longa, 1-lJ poll, lata, basi acuta, ssepe breviter in petiolum decurrentia, 
caudato-acuminata, subtus minutissime punctato-furfuracea, ^-nervia, nervis supra 
leviter impressis, subtus prominentibus, venis transversis vix conspicuis. Cymae 
axillares, pauciflorse, pier unique umbelliformes, rarius racemose disposita;, pedun- 
culo ^-1 poll, longo suffultse. Pedicelh 4-6 lin. longi. Calyx breviter cupulari- 
campanulatus, 2| lin. longus, glaber, limbo integerrimo. Petala 5 alba, late 
obovata, 4 lin. longa, obtusa. Stamina 10, sequalia ; antherse longitudine tantum 
paulo insequales, lineari-lanceolatse, acuminatae, 2-3 lin. longse ; connectivum antice 
in lobos 2 obtusos productum, postice calcaratum. Ovarium usque ad verticem 
calyci adnatum, lamellulis 10 angustis ad annulum staminigerum demum incras- 
satum ascendentibus. Bacca alba, globosa, 3 lin. dimetiens, limbo persistente 
coronata. 
Penokok, alt. 3500 feet {Haciland, 1278). 



DE. 0. STAFF ON THE FLOEA OF MOUNT KIXABALU. 161 

Allied to Medinilla crassineri-ia, Blume, and M. quintuplinervky Cogn., but distinguished 
from both by the ternate leaves, and, besides, from tlie former by the caudate apex of 
the leaves and the smaller flowers and fruits, and from the latter by the distinct 
8-nerved leaves, the 5-merous flowers, and the longer inflorescences. It approaches 
still more to an undescribed species, collected by Forbes in New Guinea (no. 932), 
which has also ternate caudate leaves, but much larger flowers. 

Medinilla lasioclados, Stapf, n. sp. Prutex. Eiimuli novelli teretes vel subtetragoni, 
dense setulis rufo-tomentosis vestiti, deraum glabrescentes ; rami vetustiores robusti, 
tetragoni vel subteretes, glabrati, lenticellis tuberculiformibus obsiti. Folia oppo- 
sita, elliptico-oblonga, 2i-3 poll, longa, \\ poll, lata, basi rotundata, apice acuta, 
crassiuscula, in gemmd suhtus demisslme rufo-stellato-tomeutosa, mox glabrescentia, 
S-nerria vel 3plo-7wrv?a, evenia vel subevenia. Cymuljc axillares, Sicpe iu ramis 
annotinis paucifiorse, J-f poll, longae, bracteatae, furturaceo-tomentellaj. Hractea; 
ovatae, |-1^ Un. longcB, acutcB^ membrcuiacece, hhnc, sub calyce persistentes. Pedicelli 
brevissimi, ^-| lin. longi. Calyx ovoideo-campanulatus, vix 2 lin. longus, rufo- 
furfiiraceus, demum glabrescens et minute bullato-rugosus, limbo integerrimo. 
Petala rosea, 4, late obovata, 2^ lin. longa, ol)tusa. Stamina 8, subuiqualia; antherae 
lineari-lanceolatse, 1 lin. paulo longiorcs, basi brevissime bilobie, lol)is conuectivi 
processubus adnatis carinatis, postice breviter calcaratae. Ovarium ad medium calyci 
adnatum, superne lamellis ad annulum staminigerum demum iacrassatum ascen- 
dentibus cum eo conjunctum. 
At 6000 feet (Raviland, 1225). 
Allied to an undescribed species from New Guinea (Forbes, 305) and to 31. rhodo- 

ehlwna, A. Gray, from the Fiji and Samoan Islands, but distinguished by tri-nerved or 

tripli-nerved leaves, much smaller bracts, and, so far as Forbes 's plant is concerned, by a 

different indumentum. 

Axplectkum homceandrum, Stapf, n. sp. Frutex bumilis. Hamuli novelli teretes, 
graciles, minute rufo-furf uracei, mox glabrati ; rami vetustiores cortice badio lougi- 
tudinaliter fisso et secundum margiaibus laciniarum soluto. Folia opposita, lamina 
in piano mediano expansa, petiolo 1 lin. longo suffulta, ovata (sine acumine), 1^1| 
poll, longa, |-1 poll, lata, basi rotundata, longe caudato-acum'mata, acuimne lineari 
\-ipoll. longo, vix I lin. lato, obtmo, coriacea, in gemma minute furfuracea, mox 
glabrata, supra opaca, subtus subnitida, Zplo-nervia vel pare infimo tenuissimo 
addito sub 5plo-nervia, nervis supra impressis, venis inconspicuis. Cymulae 3-1- 
JlorcB, breves, pedunculo 1-2 lin. longo sufultce, rufo-furf uraceaj, bracteata^. Bracteaj 
lanceolatse, acutjfi, i-| lin. longae. Pedicelli 1-2 lin. longi. Calyx breviter ovoideo- 
campanulatus, li lin. longus, minute furfuraceus ; limbus basi- circumcirca con- 
strictus, quadrilobus; lobi e basi ovata subulati, i lin. longi. Petala 4, rosea, 
oblongo-lanceolata, 2 J-3 lin. longa. Stamina 8, subceqiicdia ; antherae 4 oblongo- 
lineares, basi brevissime bilobse, connectivo non vel vix bicalloso, postice in calcar 
applanatum, obtriangulare, filamentum subamplectens producto, fertiles ; antherae 4 



162 DE. 0. STAFF 0^ THE FLOEA OF MOUNT KINABALU. 

vix minores, connectivo antice ia processus binos breves obtusos producto, postice 
in ealcar breve conicum abeunte. Ovarium ad verticem subplanum calycis tubo 
adiiatum. Bacca globosa, 1^ lin. dimetiens, calyce persistente coronata, rubra. 
At 6000 feet (Havilcmd, 1170). 

Closely allied to Anplectrum myrtiforme, Naudin (=Medinilla myrtiformis, Triana), and 
distinguished only by the longer acumen of the rather tripli-nerved than quintupli-nerved 
leaves and by much shorter inflorescences. This species and likewise A. myrtiforme are, 
in spite of their almost equal anthers, true Anplectrums . They have quite the habit of 
Anplectrum^ and agree also in the structure of the leaves and of the bark. Whether the 
smaller set of the anthers is fertile or not, I am not able to m.ake out from the material 
at my disposal. 

KiBESSiA TESSELLATA, Stapf, n. sp. E,ami obtuse tetragoni, cmgulis anguste crispo-imdu- 
latis, glabri. Folia sessilici vel subsessilia, ovata vel ovato-oblonga, 3J-5|^ poll, longa, 
2-2J poll, lata, basi late rotundata, plerumque plus minusve cordata, apice sensim 
in acumen angustum obtusum producta^ tenuiter coriacea, concoloria, glaberrima, 
5-nervia, nervis supra impressis, infra cum venis transversis prominulis. Cymulae 
3-l-flor8e, axillares, singulse vel paucae fascicnlatge, ^-1 poll, longae, bracteatse, 
glabrae. Bractea? ovatse, minutse. Calycis tubus hemisphaericus, squamis rhomhoideis, 
crassis, suhplanis vel mnhonatis, vel summis ovatis acutis vestitus, sub anthesi 2 J-3 lin. 
longus ; limbus late conicus, caudato-apiculatus, 3 lin. longus. Petala lilacino- 
cserulea, ovato-triangularia, acuta, 3 lin. longa. Anthera3 dolabriformcs, utrinquc? 
inappendiculatge, 1 lin. longse. Bacca depresso-globosa, 5 lin. dimetiens, squamis — 
marginalibus exceptis — applanatis, inconspicuis. 
At Koung, on the Tampassuk Biver, alt. 2000 feet {Haviland, 1303). 
Closely allied to K. simpjlex, Korth. {K. echinata, Cogn.), and to an undescribed species 

from Penang ( C. Curtis, 953), but differing from both in the winged stem and the sessile 

leaves, and, besides, from the former in the glabrous and short scales on the calyx and 

from the latter in the gradually tapering apex of the leaves. 

Memectlon l.evigatum, Blume, Mus. Bot. i. 358. 

Low flooded land, near Tawaran {Haviland, 1384). 

Distribution : Malaya, on the mainland as far as Tavoy. 

The distribution of this species is in all probability much wider, as the characters 
supposed to distinguish it from several of the allied species are so trifling and so artificial 
that they can hardly be considered sufiicient to maintain it as a distinct species. It 
approaches so near to M. edule, Boxb., that either part of that species must be transferred 
to M. IcBvigatum, or 31. Icevigatmn should be included in it as a variety. In any case, 
the species of Memecylon have been multiplied far too much ; a revision, however, of 
the genus, based on dried material only, would be very difiicult. 

Memecylon lanceolatum, Cogn. in DC. Monogr. vii. 1157. 
At 6000 feet {Haviland, 1168). 
I do not consider this determination as definitive. The species belongs to the doul)tful 



DR. 0. STAFF ON THE FLORA OF MOUxXT KIXABALU. 163 

group just mentioned. It difFers f rom typical Ilemeeylon la>i'igatum chiefly in the narrower 
leaves with a longer acumen, in the longer pedicels, and a more conspicuous, distinctly 
4-lobed calyx-limb. The leaves of Jf. lanceolatum vary from lanceolate (12-2 : h in.) 
with hardly any acumen to elliptic-oblong (l|-2 : 1 in.) with a very distinct acumen, 
and are 4-6 lin. in length. It is the latter form (Beccari, S. P. 2424) wliich is represented 
in the Kinabalu specimen, whilst the former (Beccari, 2118) is tlie type from which 
Cogniaux drew his description. The leaves of Dr. Haviland's specimen are olive-green to 
olive-brown above and yellowish green beneath, whilst those of Beccari's specimens are 
very dark above and olive-brown beneath, and an analogous dilferencc is obvious in the 
colouring of the young branchlets. I doubt, however, very much whether this difference 
is a character of specific value. The berries of tlie Kinabalu plant are smooth, 
about 4 lin. in diameter, and supported by a pedicel 1-li lin. long. The pedicels rise 
from a very short common peduncle, and bear each two minute bracts at the base. 

Plethiandra Hookeri, Stapf, n. sp. Frutex. Rami crassiusculi, teretes, cortice 
cinereo-albido. Folia petiolo 1 lin. longo, crasso suffulta, vel subsessilia, 2^-4 poll, 
longa, 1^-2 poll, lata, basi subrotundata, apice obtusissima vel subretusa, crasse 
coriacea, subconcoloria, utrinque nitidula, 5pli-nervia, paribus nervorum remotis, 
evenia. Flores axillares 3, rarius 4 vel 5 fasciculati. Pedicelli 2-3 lin. longi, 
crassiusculi, purpurei. Calyx urceolato-campanulatus, 2J lin. longus, glaberrimus, 
purpureus, limbo crassiusculo integro. Petala rosea, obovato-elliptica, acuta, 3 
lin. longa. Stamina vix 30 annulo incrassato inserta, exteriora 6 vel paulo plura 
longiora; filamenta circa 2 lin. longa; antherae lineares, obtusae, 1 lin. longa}, hasi 
inconspicue bilohce, postice minutissime calcarafw, apice poro dehiscentes. Ovarium, 
usque ad verticem in medio leviter depressum, calyci adnatum, Q-locidare ; septa 
tenuissima ; placentae bifidsB ex angulo superiore ortae. Bacca (immatura) globosa, 
limbo calycino coronata, in vertice annulo staminigero persistente instructa. 
At 6600 feet (Ilcwilcmd, 1169). 

There is no doubt that the plant just described belongs to Plethlaadra, a genus founded 
by Sir Joseph Hooker on a rather poor specimen collected by Motley in Labuan. This 
genus was placed at the end of Melastomaceae by the authors of ' Genera Plantarum,' 
and amongst Astroniese by Cogniaux. The excellent material which is now in my hands 
enables me to determine its true position. It is nearest allied to Jledinilla — so near, 
indeed, that it could be treated almost as a polyandrous section or subgenus of it. The 
general habit is exactly that of Medinilla, and the floral structure diff'ers but little, 
apart from the number of the stamens. There are 6-merous species in various groups of 
Medinilla. The stamens are inserted, as in Medinilla, on the thickened annular internal 
margin of the calyx-tube in a single series, or nearly so ; but the anthers are grouped 
finally more or less in concentric rows. Their shape differs from that of typical Medinilla 
in the almost absolute suppression of the processes of the connective and in their exactly 
linear shape. They open wdth pores like Medinilla. The authors of ' Genera Plantarum' 
say " rimis longitudinalibus antice dehiscentibus." I have not observed this mode of 
dehiscence either in P. Eookeri or in another new species from Sarawak, collected by 



164 DR. 0. STAFF ON THE FLORA OF MOUNT KINABALU. 

Dr. Hayiland ; and I find it only in a single flower in the type specimen. In tins 
the few anthers which are still attached are slit from the top to the base, but the margins 
of the cell gape so much that the cell-septa are exposed, and, as the margins are more- 
over jagged, I suppose the state of these anthers is abnormal. On the other hand, we 
know certain MecUnillas the anthers of which open with a short terminal slit. There- 
fore, if the anthers of P. Motleyi, Hook, fil., should indeed always open with a longi- 
tudinal slit, this would in no degree disprove the affinity of metUandra and 
Medinilla. The ovary is described by Sir Joseph Hooker as 4-locular. The only ovary of 
P. Motleyi which I dissected was in a rather advanced state, when the exceedingly thin 
septa are very apt to break either by the expansion of the growing fruit or by the 
shghtest pressure. But, if I was not very much mistaken, it is also 6-merous. The very 
delicate structure of the septa may easily account for the oversight of one or more of 
them. The placentation is also exactly as in lledinilla. 

SAMYHACE^. 

Casearia leijcolepis, Turcz. in Bull. Soc. Nat. Mosc. (1858) i. 463. 
On the Tawaran Biver {Haviland, 1388). 
Distribution : Malaya, from Java to the Philippines. 

BEGONIACE^. 

Begonia adenostegia, Stapf, n. sp. {^ Platy centrum) . Herba epiphytica. Caulis 
gracilis, inferne induratus atque radicans, superne rubellus vel ruber, breviter cris- 
pule hirsutus. Eolia petiolo, rubello crispule hirsuto, 3-5 lin. longo suffulta, valde 
oblique ovata, 3-4 poll, longa, IJ-lf poll, lata, basi valde asymmetrica, unilateraliter 
cordata, in latere extus spectante in lobum magnum rotundum producta, in latere 
interiore non vel vix auriculata, apice acuminata, in margine obscure denticulata, 
firme membranacea, supra glaberrima, subtus in nervis adpresse rufo-furfuraceo- 
hirsuta, in parte exteriore nervis basalibus 3-4, altius 2-3 lateralibus additis, in 
interiore nervo solitario basali percursa, venulis imprimis supra prominulis reticu- 
lata. Stipulse ovatae vel oblongae, 3-4 lin. longae, acutee vel obtusae, margine plus 
niimisve glcmduUs brevissime crasseque stipitatls obsitce, demum deciduge. Inflores- 
centia terminaUs, rariits pseudolateralis et folio opposita ; masciila racemosa vel 
puree paniculata, plurifloray lJ-3 ^wll. longa ; feminea tmiflora^ 1 J poll, longa ; m 
uti^oque sexti pluribracteata. Bracteae oppositae, cordato-ovatae vel cordato-rotun- 
datae, inferiores 3 lin. longae, membranaceae, venosae, breviter glandulose fimbriatm. 
^ Pedicelli florum masculorum 2-1 lin. longi, femineorum demum 6-8 lin. lon^^i. Elos 

masculus : Sepala alba rotundata, basi connata, 3 lin. longa. Petala alba 2, oblonga, 
2 lin. longa. Stamina monadelpha ; filamenta ima basi connata, -J-| lin. longa ; 
antherae ovatae vel oblongae, apice emarginatse. Elos femineus : Perigonii lobi 5, 
albi, ovati, majores 4 lin. longi. Ovarium glabrum, trialatiim, alis oblique obovatis 
in pedicello breviter decurrentibus ; styli 2, 1 lin. longi, iteratim dichotomi ; 
stigmata in divisionum apicibus capitata, valde papillosa ; loculi 2 ; placentae bifid^e. 



DR. 0. STAFF 0^ THE FLOEA OF MOUNT KIXABALU. 165 

Capsula nutans, 6 lin. longa, alis 3 inaequalibus, luajore subquadrato-rotundata, 
6 lin. lata, minoribus oblique oboyatis, 3^ lin. latis, facie ventrali (alis minoribus 
circumdata) circumscissa, decidua, faciebus dorsalibus priino juxta alas minores, 
demum circumcirca dehiscentibus. 

Kinitaki, alt. 4500 feet (Hamland, 1270). 

Allied to Begonia varicms, A. DC, and B. rupicola, Miq., but very distinct by the 
peculiar inflorescence and the stalked glands of the bracts and stipules. 

Begonia Burbidgei, Stapf, n. sp. (§ Bmcteibegonia). Herba G-8 ped. alta. Caulis 
inferne induratus, superne parce crispule liirsutus, internodiis ^-IJ poll, longis. 
Polia petiolo parce liirsutulo purpui-ascente \~\\ poll, lonji^o suffulta, vahlc 
asymmetrica, oblique cordato-ovata, 3-4 poll. longa, 2 poll, lata, in latere 
exteriore in lobum ampluin rotunduni ])roducta, in latere interiore inauri- 
culata, apice acuminata, repando-denticulata, firmula, utrinquc glaberrima, 
lucida, in parte niajore nervis basalibus 4-5 atque lateralibus 2-3 additis 
(intermediis 3 eximie bifurcatis), in parte minora nervis l)asalil)us 2, lateralibus 
2 additis, omnibus purpurascentibus, venulis inconspicuis. Inllorescentia niaaaiUa 
tennlualls, 1-2-pollicaris, imrce dichotome ramosa, '3~o-/lora, hdeniodus (infimis 
interdum exceptis) hractels multo brerioribus ; feminea ex folii summi axilld uniflora 
vel dichotoma bijlora^ 1^-2 poll, longa, bibracteata. Bracteic in inflorescentiA. 
mascula opposite^, imbricatce, late obovatce, 5-S lin. lougcu^ membra?iace(Uy viridi- 
albidce, glaberrimce, j^ersistentes. Pedicelli 2 lin. 1 »n§} ^el subnulli. Flos masculus : 
Sepala 2 rotundato-quadrata, 6-10 lin. longa, albida, roseo-suffusa. Petala 2, ovato- 
oblonga, 4 lin. longa, alba. Stamina basi connata; tilanuT.ta ^-1 lin. longa; 
antherse lineari-oblongye, obtusae, 1 lin. longae. Flos femineus : Perianthii lobi 5, 
late obovati vel oblongi, majores ad 9-15 lin. longi. Ovarium subfequaliter triala- 
tum, glaberrimum; styli 3 breviter bicrures; stigmata in marginibus crurum tort- 
orum ; loculi 3 ; placentae bifida?. Capsula a basi juxta alas dehiscens. Semina 
oblonga, ^ lin. longa, aj^iculata, laxissime reticulata. 

At 7600 feet [Haviland, 1188) ; in wet shady jungle above Kiau (Biirbidge). 

A very remarkable species, distinguished by the large, crowded, and persistent 
bracts. 

Begonia oblongifolia, Stapf, n. sp. [^ Betermannia). Caulis breviter crispule liir- 
sutus. Folia petiolo 4-2 lin. longo eodem indumento ac caulis vestito sutfulta, 
oblique oblonga, 4-6J poll, longa, 1| poll, lata, in latere exteriore majore in lobum 
rotundatum brevissimum petiolo adnatum producta, in interiore rotandata, apice 
acuminata, serrulata, tenuiter membranacea, supra obscure viridia, ima basi excepta 
glaberrima, subtus pallidiora, in nervis f urfuraceo-puberula, nervis basalibus in latere 
exteriore 2-3, in latere interiore 1, lateraliljus utrinque 6-T angulo vix 40° ortis, 
subrectis. Stipulae lineari-oblonga?, 6-7 lin . iongse, acuta?, membranaceae, f urf uraceo- 
puberulse. Inflorescentia (mascula tantum nota) terminalis, pauiculata, multi/lora, 
furfuraceo-puberula, bracteata, fere 3 poll, longa. Bractese sub ramis primariis 
stipulis conformes, paulo minores, cseterum minimge. Pedicelli graciles, ad 2 lin. 

SECOND series. BOrANT, VOL. IV. Z 



166 DE. 0. STAPP ON THE FLOEA OF MOUNT KINABALTJ. 

longi. Elos masculus : Sepala alba, rotunda, vix 2 lin. dimetientia. Petala nulla ; 
filameiita basi connata brevissima ; antherae obovato-cuneatae, obtusae, vix J lin. 
longae. Capsula (ex icone) obovato-oblonga, iiltra-pollicaris, trialata, alls aequalibus, 
2 lin. longis. 
Near the Dahombang River, alt. 3000 feet {Haviland, 1308). 

Very like Begonia lepida, Blume, in general habit, but easily distinguished by the 
profuse inflorescence and the small apetal flowers. The latter character brings it to the 
section Fetermannia, which is separable only artificially from § Bmcteihegonia. 

Begonia borneensis, A.DC, Prodr. xv. part i. 320 (§ Petermannia). 

At 6000 feet (^rty^7a/?c/, 1189). 
Distribution: Borneo. 

There is no description of the male flowers in De Candolle's ' Prodromus.' Therefore 
I give it here : — 

Infiorescentki mascula terminalis, interdum ex folii summi axilla altera addita, panicu- 

lata, ramulis gracilibns, longe racemosis, ad IJ poll, longis, bracteatis ; bractea? 

oblongse, caducse, IJ lin. longse vel minores, membranacese. Sepala alba, ovata vel 

oblonga, 1 J lin. longa. Petala nulla. Antherce in toro elongato subsessiles, obovatte, 

^ lin. longge. 

The female inflorescence is axillary, from the axil of the uppermost leaf. It is reduced 

to a short peduncle with two pedicels about 9 lin. long. In its place there is in one of 

the Kinabalu specimens a male inflorescence. The capsules of the type are a little 

longer and their pedicels shorter, but the difference does not seem to me sufficient to 

make the Kinabalu plant a variety. 

Begonia inostegia, Stapf, in Hook. Ic. PI. t. 2309 ' %Fetermannia) , Caulis, ut petioli et in- 
florescentia, breviter crispule hirsutus, inferne glabratus. Folia petiolo 3-1 poll, longo 
suffiilta, late et oblique cordato-ovata, 5-6 poll, longa, 4-5 poll, lata, sinu latissimo, 
apice acuminata, repando-denticulata,tenuiter membranacea, subconcoloria,maculata, 
supra basi excepta glaberrima, subtus in nervis minute furfuraceo-puberula, in parte 
majore exteriore nervis basalibus 5-6, in minore 3, lateralibus 3 vel 2 additis, fere 
omnibus semel vel bis furcatis. Stipulae persistentes, ovatce acutce, 4-7 lin, longae, 
nervis crebris in setas 4-7 lin. longas cxcurventihus ci'inilo-jimhriatce. Inflorescentia 
mascula terminalis, iteratim tri- vel dichotoma, pedunculo 10 lin. longo suffulta, f 
poll, longa, in racemulos breves abiens, bracteata. Bracteae stipidis conformes, 
minores siimmce ohlongce^liaudjimbriatai. Pedicelli 2-1 lin. longi, graciles. Inflores- 
centia feminea ex folii summi axilla ad flores geminos ebracteatos redacta, pedi- 
cellis recurvis 10 lin. longis. Plos masculus : Sepala 2 alba, ovata vel oblonga, 
2 lin. longa. Petala nulla. Stamina in toro filiformi elongato inserta; filamenta 
brevia vel brevissima ; antherie obovatse \ lin. longae. Flos femineus ignotus. 
Capsula obovoidea, 5-6 lin. longa, alls 3-lJ lin. latis basi ultra loculos productis ; 
loculi 3 juxta alas dehiscentes; placentae integrie. Semina minuta, breviter cylin- 
drica, \ lin. longa, laxissime reticulata. 



DE. 0. STAFF OX THE PLOKA OF MOUNT KIXABALU. 107 

At 6000 feet (Hamland, 1190). 

Very remarkable from the long fibres into wliicli the nerves of tlie stipules and bracts 
run out. Otherwise the plant approaclies Begonia hornrcnsis, A. DC, in tlie s^eneral 
habit and in the floral structure. 

UMBELLIFER.E. 

Hydrocotyle javanica, Thunb. Diss. ii. 415, t. 2. 

Bed of the Dahombang, alt. 3000 feet (Uiwilaud, 1273). 

Distribution : All over Malaya to the Pliilippines and Australia ; Indo-China to Ilupeh 
and Ichaiig" ; Khasia, Himalaya ; Ceylon and jNIalabar. 

According to Dr. Haviland, the 2)lant is said to be used by tlie Dusuns like *' tuba," 
for poisoning fisli. 

The specimen referred to H. javanica in Fl. of Troj). Africa iii. 4 is iiot tliat species. 

Teachymene sanicul^folia, Stapf, in Ilook. Tc. PI. t. 2308. Perennis. Rhizoma ])ro- 
stratum vel ascendens, ramosum. Rami in individuis in rupium fissuris crescentibus 
valde abbreviati, dense foliorum residuis squamiformibus vcstiti, sub ips;X facie rupium 
iteratim ramulosi, ramulis brevissimis dcnsissime foliis vestitis et scnpos unibc'lligcros 
edentibus ; in individuis soli liumidioris sappe valde elongati, in parte subter- 
ranea crebre radicantes, in parte supraterraneA, e basi longe prostratA ascondentes, 
pedunculos ex axillis foliorum edentes, caulis badii patule rufo-hirsuti inteniodiis 
in planta luxuriante ad 1 poll, longis vel longioribus. Folia petiolo glabro vel riifo- 
hirsuto 2J-4 poll, longo sufFulta, ambitu cordato-orbicularia, plerumque transverse 
latiora, ad § triloba, lobis obovato-cuneatis, acute tridentatis vel in majoribus trilo- 
bulatis, lobulis tridentatis et dentibus interdum in latere exteriore denticulo auctis. 
4-12 lin. longa, 8-18 lin. lata, crassiuscula, glaberrima vel utrinque hirsuta. 
Pedunculi (vel scapi) teretes, nudi, 2^-9 poll, longi, tenuiter striati, rubelli. Invo- 
lucri phylla linearia vel lanceolato-linearia acuta, 2-4 lin. longa, glaberrima vel 
hirsuta ciliataque. Pedicelli sub anthesi phyllis a^quilongi, deinde paulo longiores, 
glabri vel parce hirsutuli. Calycis dentes persistentes, triangulares, distincti. 
Petala alba, obovata vel elliptica, 1-1 j lin. longa, supra costula mediana notata. 
Anther a? rosea?, ellipticie. Styli demum deflexi, vix \ lin. longi. Fructus a latere 
valde compressus, reniformis, l^-lj lin- longus, 1|-2J lin. latus; mericarpia 
Eequalia vel subaequalia, glaberrima, atropurpurea vel brunnea, exalata, jugis inter- 
mediis tenuissimis, marginibus interioribus approximatis. 
From 7000 feet to the top, in the higher region in crevices of the rock and very 
stunted (Loic, Burhidge ; Haviland, 1130, 1162). 

Nearest allied to T. Immilis, Hook. fiL, from the alpine and subalpine region of Victoria 
and Tasmania. The stunted, glabrous form agrees perfectly with T. humilis so far as its 
growth is concerned. The involucrum, the umbels, the flowers and fruits are, apart 
from the hairiness, exactly the same in both of the Kinabalu forms, and they differ hardly 
at all from those of T. humilis. Thus the main diff'ereuce is in the shape of the leaves, 
and is rather considerable. The stunted and the creeping forms from Kinabalu look at 

z2 



168 DE. 0. STAFF ON THE PLOEA OF MOU^T KIISABALXI. 

the first glance very different, but this difference is evidently and exclusively due to the 
extreme abbreviation of the internodes in the stunted form, and this again is caused by 
the conditions of the habitat. The hairiness or glabrousness is still less constant. There 
are hairy and glabrous stunted, and hairy and glabrous prostrate specimens. The same 
kind of hairiness, though in a minor degree, may be found on the petioles and the 
involucre of a specimen of Trachymene humilis from the Limes^^^one River, collected by 
Sir F. V. Mueller, whilst all other specimens I have seen are glabrous. 

BUBIACEiG. 

AiiTHROPHYLLUM DivERSiFOLiuM, Blume, Bijdr. 879. 

On the Tampassuk River near Kiau, alt. 3500 feet {Havilandy 1254). 
Distribution : All over Malaya, from tlie Andamans and Penang to North Borneo. 

Wendlandia paxiculata, DC, Prodr. iv. 411. 

In young jungle near Kiau, alt. 3000 feet {Haviland, 1366). 

Distribution : Malayan Archipelago, from Java to Borneo and the Moluccas. 

AiiG.osTEMMA BEACHYANTHEuuM, Stapf, u. sp. Herba simplex e basi prostrata ascendens, 
circa 4 poll. alta. Caulis adpresse hirtello-pubescens. Polia opposita, valde ina?qualia ; 
majora petiolo 1-lJ lin. longo suffulta, elliptica vel ovato-elliptica vel obovato- 
elliptica, IJ poll, longa, 7-9 lin. lata, basi rolundata et plerumque plus minusve 
asymmetrica, breviter acuminata, membranacoa, supra obscure viridia, glaberrima 
vel marginem versus vel in margine parce pilosula, supra brunnea, in ncrvis adpresse 
hirtello-pubescentia, nervis utrinque circiter 8: minora sessilia, plerumque ad subulam 
brevem redacta, rarius lamina ovato-lanceolata ad 3 lin. longa. Stipulse ovatae ad 
3 lin. longse, obtusse, minute et parce hirtellc3e, tenuiter membranacete. Inflorescentia 
terminalis, breviter pedunculata, umbellatim 3-flora vel ad florem solitarium redacta ; 
pedicelli hirtello-pubescentes 3-4 lin. Ion gi. Calyx turbinatus, adpresse hirtellus; 
tubus 1 lin. longus, dentes triangulares 1 lin. longi. Corolla alba, breviter campa- 
nulata, 2 lin. longa, ad medium h-lohata, extus densiuscule hirtella, intus in lobis 
ciliatis papillosa. Eilamenta brevissima, tubo imd basi inserla, antlierce lemter 
coh(Brente8, ohlongcB, J lin. longcB, breviter obtuseque a^iculatce, rima longitudinaliter 
dehiscentes. Stylus filiformis, 2 lin. longus ; stigma capitatum. 

At 3000 feet {Raviland, 1326). 

Quite of the hal^it of ^. tiniflorum, Blume, but very different in the structure of the 
andra^cium ; besides, the lobes of the calyx and the corolla are shorter and broader. 

AiiGosTEMMA GRAciEE, Stapf, n. sp. (Pi. XII. D. 10-12.) Herba prostrata in truncis 
arborum deciduarum. Caulis radicans, glaber. Folia opposita, valde inscqualia, 
insigmter disticha ; majora brevissime petiolata, oblonga, 5-7 lin. longa, 2-2^ lin. 
• lata, basi rotundata, apice apiculata, membranacea, supra obscure viridia, subtus 
pallida, utrinque in costa fulvo-hirtella, evenia ; minora subsessilia, ovata, ad 3 lin. 
longa, basi in latere extus spectante auriculata lobo acutiusculo. Stipulge foliacese, 



PE. O. STAFF OX THE PLOEA OF MOUNT KINABALU. 169 

breviter cordatae, i-| lin. longoe, minute setaceo-caudatoe. Flos solitarius, terminalis, 
subsessilis. Calyx campanulato-turbinatiis ; tubus adprcssc birsutus, If lin. 
longus ; lohl wquilongi, late ovati, ohtusinscnlL Corolla alba, late canipaimlata, 
4 lin. longa, fere ad medium 5-loba, glabernma ; lobi ovati acuti. Stamina 5, in 
ima basi corollte inserta, libera ; Jilmnenta nigricantia, \ I'm. longa ; antherce 
oblougce exapiculatcB, longUudinaliter dehiscentes, \ lin. longw. Cai)sula liemi- 
spba^rica, 2 lin. longa, lobis calycinis paulo auctis coronata. Semina numerosa, 
angulata, J lin. longa. 
On rotting fallen ti'unks of trees, at the Penokok, alt. 3500 feet {Ilaciland, 1325). 
In habit very like A. parvifolium, Benn., but distinguished by the short free stamens, 
and by the much shorter and broader calyx lobes and the glabrous corolla. I suspect 
that some or all species of this interesting genus have dimorphic flowers. 

Hedyotis protrusa, Stapf, n. sp. (§ Diplophragma). (PI. XII. A. 1-3.) Fruticulus 
(an herba lignescens ?). Rami stricti, obtuse tetragoni, superne brevissime tomentelli, 
nigrescentes, inferne glabrati, pallesccntes. Folia sessilia, oblongo-lanceolata, ^i poll, 
longa, Ig -2 lin. lata, mepe plicata^ basi ct apice acuta, crasaluscidc coriacea, rigida, 
glaberrima^ evenia. Stipule triangulares, 1 lin. longa*, obtusa?, basi firraa), superne 
et in margine albo-membranacea^, demuin lacera). InflorescentiaLi axillarc8f capi- 
tatim contractce, pedunculo 1-2 {raiHiis 4) tin. longo suffultcCt plerumque hracteia 
binis foliaceis capitulmn subaquanfibus instructcc. Pedicelli subnulli. Caly 
glaberrimus, nigrescens ; tubus breviter turbinatus, -^-| lin. longus; lobi 4 lineares 
obtusi, 1 lin. longi, crassiusculi, siuu latiusculo distantes. Corolla alba, in fauce et 
in tubo 1-1 4 lin. longo intus pilosa; lobi 4 ovato-oblongi, gequilongi, obtusiusculi. 
Stamina exserta ; filamenta | lin. longa, antherai \ lin. longte. Stylus 1^ lin. longus ; 
stigma breviter bilobum. Capsula ad denies medlos protnisa^ in coccos binos secedens, 
1 lin. longa ; cocci facie fere ad basin atque in vertice loculicide dehiscentes. 
Semina numerosa, angulata, nigra, J lin. longa. 
Maripari Spur, alt. 5500 feet (Haviland, 1265). 

Nearest allied to an undescribed species from Manilla in the Philippines (Voy. de la 
Bonite, no. 214?). But in this species the axillary inflorescences are very loose and 
almost 3 in. long, and the top of the capsule does not protrude beyond the calyx-tube. 
Tliere is some afSnity with several of tlie Ceylonese and Western Peninsular species 
which group round H. stylosa, Br., and it is, though remote, perhaps more evident 
than with any other species I know. 

Hedyotis pulchella, Stapf, n. sp. (§ Diplophragina). (PL XII. b. 4-6.) Bami 
superne obtuse tetragoni, glabri. Folia petiolo 2-3 fln. lon^o suffulta, elliptica vel 
ovata, 2-2^ poll, longa, l-lj poll, lata, basi acuta, saepe paulo asymmetrica, apice 
acuminata, membranacea, utrinque striolis albidis notata, demam glabrata, supra 
olivaceo-viridia, subtus Isete viridia, nereis lateralibus tdrinque 7-8, inferioribus 5-6 
approximatis, vix 2 lin, inter se distantibus, cseteris remotis. StipuljB ovato-tri- 
angulares, 3 lin. longse, laciniis angustissime linearibus hirtellis fimbriatse, albo- 
striolatse. Inflorescentise axillares, corymboso-i^aniculatse, laxae, graciles, pedunculo 



170 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

gracili li-2| poll, longo suffultic, ad ramificationes foliaceo-bracteatse, glabeiTimoc, 
albo-striolat^e. Bractese lanceolatge vel superiores lineari-lanceolatse. Pedicelli 
3-1 lin. longi. Calyx laete viridis, glaberrimus ; tubus turbinatus, 1-lJ lin. longus ; 
lobi 4, linear i-lcmceolaU, l-|-2 lin. longi, longe aciitati. Corolla alba breviter 
infundibuliformis, extus glabra ; tubus vix 2 I'm. longus, intus pilosus ; lobi 4, ovato- 
oblongi, subacuti, sequilongi, iutus a fauce albo-villosi. Stamina 4 ; lilamenta in 
fauce inserta, 1 lin. longa ; anthcrae f lin. longae, albas. Stylus 3 lin. longus; stigma 
brevissime bilobum. Capsula calycis faucem non superans, 1 tin. longa, in coccos 
2 fissa ; cocci facie late aperti. Semina numerosa, angulata \-^ lin. looga, minute 
granulata, nigra. 
From 7000 to 9000 feet (ITamland, 1122). 

Nearest allied to an undescribed species from the Philippines (Cuming, 937; Vidal, 377, 
800). This differs chiefly in smaller flowers and capsules, shorter calyx-teeth, and 
fewer nerves. It grows on Mt. Banahao, at 6000 feet. Hedyotis pulchella approaclies 
otherwise, though in a much more remote degree, R. tenuipes, HemsL, from South China, 
and H. macrostemon. Hook, et Arn., also from South China. Amongst the Indian species 
H. pmdnosa, Wight. & Arn., comes nearest; but this affinity also is by no means close. 

Hedyotis macrostegia, Stapf, n. sp. (§ Diplop)'hragma). (PI. XII. c. 7-9.) Suffrutex, 

2-3 ped. altus. Bami crassiusculi fusco-tomentelli. Polia petiolo glabro 3-5 lin. 

longo suffulta, oblongo-elliptica, 2 poll, longa, J-1 poll, lata, basi acuta, apice acute 

acuminata, crassiuscula, glaberrima, subconcoloria, olivaceo-viridia, nervis utrinque 

7-8, inferioribus 5 approxirnatis, IJ lin. inter se distantibus, angulo 30° ortis, supra 

impressis, infra valde prominulis. Stipulae amplye, libera), latissime ovatai, lin. 

longa?, lacinulis ciliatis timbriata), glaberrima), albo-striolatic, demum. albescentes. 

Inflorescentiae axillares, pedunculo ssepe fliexuoso, 2 poll, longo, hirsuto, demum 

glabrato suffultee; flores sessiles, in cymulis 3-floris capitato-congesti vel infimis 

distinctius pedunculati, bracteis foliaceis plus minusve involucrati. Bractea; 

infimge ovatse, subacuminatse, glabrae, more foliorum stipulatic, capitulo paulo 

breviores ; superiores minores, ciliatse, et supra quoque plus minusve pilosaB, basi 

utrinque lacinia fimbriata structura stipularum auctae. Calycis tubus turbinatus, 

vix IJlin. longus, albo-striolatus ; lobi 1 (rarius 5) ovati acuti, ciliati, et supra versus 

apicem pilosi, 2 lin. longi ; lobuli lanceolati vel lineares, ciliati, duplo breviores 

interjecti. Corolla alba, breviter infundibuliformis, utrinque pilosa; tubus 1-1 J lin. 

longus ; lobi ovati, 1| lin. longi, obtusiusculi, pilosi. Stamina 4 ; fllamenta 1 J lin. 

longa; antherae lineari-oblongae, 1 lin. longae. Stylus 2J lin. longus; stigma 

incrassatum, brevissime bilobum. Capsula globosa, IJ lin. longa, baud ultra 

calycis limbum producta, in coccos duos secedens ; cocci facie et vertice late aperti. 

Semina circa 10-15 in utroque cocco, angulata, -^-J lin. longa, nigra, minutissime 

granulata. 

Prom 8000 to 12,000 feet [Loic ; liar Hand, 1120). 

A very fine species, of the habit of II. Lessertiana, Arn. But the inflorescences are, as 
in the previous two species, axillary, the axis terminating with a bunch of leaves which 
grow out in the next season and again produce flowers from the axils only. However 



BR. O. STAFF OX THE FLORA OF MOUNT KIXABALU. 17t 

different the habit of the three species here descri])ed may be, they evidently belong to 
one natural group of species which seems to be limited to Xorth l^orneo and the 
Philippines, although parallels may be found amongst the continental species, in 
South China as well as in the Western Ghats and in Ceylon. 

Muss.EXDA FROXDOSA, Linne, Spec. PI. ed. I. p. 177. 

At Kian, alt. 2500 feet (Ilaviland, 1355). 

Distribution : Tropical Himalaya ; Western Ghats ; Malaya, from the Andamans to 
New Guinea and the Philippines. 

I refer this plant to M. frondosa with some hesitation, as it is evidently in a diseased 
state. It is nearest to the variety hirsutissima, Hook., hut the calyx-teeth arc broader than 
they generally are in the specimens 1 have seen from tlie Western Peninsula, and \\w 
only normal corolla is much smaller. All other corollas in the two specimens from Kina- 
balu are very much widened (8 lin. by 2 J lin.), the anthers are replaced by thick 
filaments, and the tube is quite glabrous inside. The style is also thickened and more 
or less deformed. The ovary, however, appears normal. There was ahvays a loose, 
crumb-like, black matter, perhaps the excrements of an insect, in the upi)er part of the 
tube. C. B. Clarke observed a very similar state in 31. frondosa v, hii'sutiasima at Canoor, 
and Hasskarl describes the same deformation from " M.froudosa, L., e. glabra, Vahl," 
from Java. The ovaries in the Kinabalu specimens, and still more in some from the 
Nilgherries, are swollen, as if fertilized, and in the latter resemble half-mat un^ ])errics. 
The case deserves investigation on the spot. F7'om the glabrousness of the tube I should 
suggest that the deformed flowers are female, and this is the more probable from tlie 
analogous case which Mr. Clarke has represented in a sketch. But then the only 
normal flower which came from an inflorescence with otherwise deformed female flowers 
would be male, to judge from analogy, whilst Burck found the dimorphism combined 
with dioecism. 

MussiENDA cocciNEA, Stapf, n. sp. Prutex, 1-6 ped. altus. Rami teretes, patuie rufo- 
hirsuti. Polia forma valde variabilia, nunc petiolo vix J poll, longo suffulta, ovata 
vel elliptica, 4-5^ poll, longa, 2-2^ poll, lata, basi breviter et abrupte contracta, acuta 
vel breviter acuminata, firmula, supra nigricantia, pube rufA, conspersa, subtus densius 
breviterque rufo-hirtella, demum plus minusve glabrescentia, nervis lateralibus 
primariis utrinque 8-9, inferioribus 1-5 lin. distantibus, venis transversis distinctis, 
nunc longius petiolata (ad 10 lin.), oblongo-lanceolata, 4J-6 poll, longa, 2 poll, lata, 
basi longe in petiolum attenuata, longius acuminata, supra minus rufescentia, in 
nervis venulisque tantum pilis minutis albidis adpresse hirtella, nervis eodem 
numero, sed inferioribus magis distantibus. Cymae sub anthesi valde contractae, 
demum laxiores, in inflorescentia terminali interdum floribunda racemose vel 
paniculatim, dispositce, rufo-hirtella? vel hirsutae, bracteata?. Bracteie ovato-lanceo- 
latie, 2-2J lin. longoe, dense rufo-tomentosa*. Stipulte bifidse, laciniis e basi tri- 
ansulari subulatis 2-3 lin. longis, rufo tomentosis. Pedicelli sub anthesi brevissimi, 
demum ad 2 lin. longi, dense rufo-hirtelli. Calyx ruf o-hirtellus ; tubus ovoideus, 2 lin. 
lon^us ; lohl deckhd, lanceolail, tubo vLv longiores, acuti. Corolla coccinea, extus 



172 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

rufo-hirtella ; tuUis tenuis, 1 j;c»//. longm, in parte tertia superiore paulo dilatatus, 
intus aureo-pilosus ; lobi ovaio'lanceolati, acuminati, 5-6 lin. longi, papillosi, in 
alabastro yalde recliiplicati. Stamiua in parte tubi dilatata breviter inserta; antberae 
lineares, 3 lin. longa?, basi angnste et longiuscnle bilobse. Stylus liliformis, 1 lin. 
longus, bifidus, ramulis vix J lin. longis. Oyarium 2-loculare; placentae ex axi ortse, 
plus minusve bifida?. Bacca obovoidea, ad 6 lin. longa, Digra vel albido-maculata, 
rnfo-birtella, septis tennissimis. Semina numerosa, angulata, J lin. longa, pallida, 
minute scrobiculato-granulata. 
On tbe Tampassuk and the Dahombang rivers, from 2000 to 3i00 feet (Loic ; Bamlcmd, 
1356, 1357). 

Distribution : North Borneo. 

Beccari, no. 456, from Sarawak, a fruiting specimen, belongs evidently to the same 
species, and probably also no. 2362, which, however, has glabrous berries. 

MusscBnda coccinea is intermediate between typical Musscenda and the more herba- 
ceous forms which are referred to Acr anther a, a genus which was reduced to Musscenda 
by Baillon (Hist, des PI. vii. 449), whilst it was maintained by K. Schumann in Eng:ler, 
Katiirl. Pflanzenfam. iv. Th. 4 Abth. 63, although chiefly on an erroneous representation 
of the structure of the ovary which was introduced first by Arnott in his original diagnosis 
of the genus (Ann. Nat. Hist. iii. 20). According to this author the ovarium is "pseudo- 
biloculare, dissepimentis duobus oppositis, vix ad medium attingentibus, placentam 
bilamellatam divaricatam ferentibus." But the fact is tbat even in the species on which 
the geniis was founded, viz. A. zeylanica, the ovary is perfectly bilocular and tbat the 
bilamellate placenta rises from the axis, just as it does in typical Musstendas. But the 
division of the placenta often reaches almost the centre, and then disintegration easily 
takes place, particularly when the transverse sections are not made with a sharp 
razor. I dissected various flowers and semi-mature fruits and always found the septa 
reaching right through. So far there is no difi'erence between Acranthera and Musscenda. 
The berry o^ A. Maingayi, Hook, fil., was stated to be 2-celled by the author himself 
in Hook. fil. PL Brit. Ind. iii. 92 ; Hemsley pointed out the 2-celled structure of the 
ovary for .^. GriffitJiii, Hook. fil. (in Hook. Ic. PL t. 1718), and I find it also in.^. tomen- 
toscf, Brown, A. gi^andiflora, Bedd., and A. cmamallica^ Bedd. Nevertheless, I think, 
Acranthera must be retained as a well-marked genus after excluding A, Maingayi and 
A. Griffithii. It may be characterized by the herbaceous growth, the generally 
elongated ovary and fruit, the blue or reddish, upwards widened and funnel-shaped or 
campanulate corolla, and the club-shaped, entire style. It is known from Ceylon, the 
Tinnevelly and Anamally Hills, the Khasia Hills, Cachar and Manipur, and there 
are several species, still undescribed, in the Kew Herbarium, from Borneo. Acranthera 
Maingayi, A. Griffithii, Musscenda mutahilis, HemsL, M. kiutaensisy King, mss. (Perak, fl. 
Mai. Pen. no. 4255, King's Coll.), and a new species from Borneo (Sarawak, LoIjIj), form 
another very natural group, which are nearer to the Indian Mussacndas ; they still difi'or, 
however, in the habit by the straggling or bushy growth, larger corollas, and the absence 
of an enlarged calyx-lobe. The species enumerated here are all very closely allied, and 
will perhaps be partly reduced in future. Their corollas are of a brilliant orange 



BR. 0. STAFF OX THE FLOKA OF 310UXT KLXABALU. 175 

colour, wliich sometimes turns scarlet. This group is connected with the true Mussffiudas 
by JI. coccinea, which has no enlarged calyx-lol)e and scarlet floAvers, but otlierwisc 
comes very near to them, particularly in the leaves and fruit. These species can hai-dly be 
separated from MusscBuda, but they might be made conveniently into a separate section, 
for which I propose the name Asenianthia. 

The leaves of M. coccinea, as it is represented by no. 1357 of Haviland, differ at the 
first glance considerably from the fruiting specimen, 135(5, and from Low's si)8cimen. 
But there is no doubt about the identity of the latter and 135G, and as the inflorescence 
and the flowers of Low's plant and of no. 1357 entirely agree, I have no doubt that 
they all belong to one species, especially as one of the leaves of no. 1357 has a much 
shorter petiole and is comparatively broader, thus approaching the shape of the leaves in 
Low's plant, and as the nervation follows the same type, though the nerves are more 
distant in accordance with the lengthened shape of tlie blade. 

Burck (Ann. du Jard. Buit. iii. 108-112) lias slioAvn that at h^ast several of the 
Musssendas are dicecious. It is rather diflicult to check his investigations on dry 
material. The structure of the corolla is certainly of the male type, and the anthers are 
polliniferous in the tAvo flowering specimens in my hands ; further, the style is very short 
and in all probability imperfect. The ovary is apparently fertile, but the numerous ovules 
did not give me tlie impression that they are so. The si)eciinen of M. mutahUis figured 
in Hook. Ic. PL t. 1718 is male also ; but the specimen of the new species, collected by 
Lobb in Sarawak, seems to be female. The corolla tube is a little widened above, but 
gradually and much less than in the male flowers of J/, coccinea and M. ftiuiabilis, and 
the anthers are indehiscent; the corolla tube, however, is golden papillose down to their 
insertion, as one would expect for male flowers. On the other hand, the style reaches to 
the tips of the anthers and divides into two subulate branches which slightly surjiass the 
mouth of the corolla. 

AcRANTHERA Atropella, Stapf, u. sp. Hcrba perennis. Caulis nigricans, sparse et 
tenuiter adpresse sericeo-pilosus, denium glabratus. Folia petiolo 1-1 1^ jwll. longo 
adpresse tenuiterque pubescente suffulta, elliptica vel elliptico-oblonga, 4-5 poll, 
longa, 1^-21 poll, lata, utrinque sequaliter acuta, membranacea, supra nigricantia, 
o-laberrima, exsiccata ob epidermatis cellulas collapsas minutissime foveolato- 
punctata, subtus fusca, in nervis venisque et liinc inde in parenchymate pilis 
tenuibus adpressis pubescentia, nervis lateralibus primariis G-8 valde prorsus 
arcuatis, venulis indistinctis. Stipulaj (summai mihi tantum notae) oblongce, |-1 
poll. longcB, obtusissimai, nigricantes, sparsissime adpresseque pilosse vel glabratae. 
Inflorescentia terminalis, pseudo-uinbellata, sessilis, foliorum paribus 2 summis valde 
approximatis suffulta et eorum stipulis subinvolucrata. Pedicelli nutantes 2 lin. 
longi, griseo-pubescentes. Calycis adpresse pubescentis tubus subcylindricus, sursum 
paululo dilatatus, vix 4 lin. longus ; lobi 5, ovato-lanceolati, circiter 4 lin. longi, roseo 
suffusi. Corolla obscure cyanea, Icfte tuhuloso-campanalata, extus griseo-pubescens 
intus glaberrima ; tubus 5 lin. longus, 2 lin. latus ; lobi 5, oblongo-lanceolati, 
subpatuli, apice recurvi, 3 lin. longi. Stamina 5, ima basi coroUoe inserta ; filamenta 

SECOND SERIES. — BOTANY, VOL. lY. 2 A 



174 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

brevissima ; antherse conniventes, primo subcohgerentes, lineares, vix 3 lin. longae ; 

connectivuni in subulam ultra loculos productum. Stylus crassiusculus, 3 lin. 

longus ; stigma capitatum, papillosum. Ovarium biloculare ; placentae ex axi ortse, 

insequaliter et crassiuscule bilamellatse, multi-ovulatge. 
On damp places, alt. 5000-6000 feet {Low; Haviland, 1237). 

This species belongs to a group wliicb is a representative of the Acranthera species of 
Ceylon, the Tinnevelly, the Anamally, and the Khasia Hills. I know three or four species 
from Borneo. They differ from the former chiefly in the short and wide corolla and the 
large upper stipules, which form a kind of involucre. The structure of the androecium 
and gynsecium, however, is essentially as in ^. zeylanica, Am. 

LuciN^A PENTACME, Stapf, n. sp. Frutex scandens, glaberrimus. Eami novelli 
tetragoni, nigricantes, annotini cortice pallescente. Folia petiolo 2-3 lin. longo 
suffulta, elUpticO' vel lanceolato-oblonga, l|-2i poll, longa, f-1 poll, lata, utrinque 
acuta vel basi subrotunda vel apice subacuminata, coriacea, flavido-viridia, partim 
exsiccando nigricantia, nervis lateralibus indistlnctis utrinque 7-8 vel inconspicuis, 
evenia. Stipula3 purpurascentes, fere ad medium in tubum connatse, ovatae, 
acimiincit(^, diutius persistentes, demum deciduse. Capitula terminalia, solitaria, 
pedunculo j-f poll, longo suffulta, basi bracteis in cupulam connatis obtusissimis 
involuerata. Calyx (sub anthesi) liber, ovoideo-cupularis, 3 lin. longus; limbus 
truncatus. Corolla roseo-albida ; tubus brevissimus, fauce albo-papilloso-villosa ; 
lobi in alabastro ad medium arete cohferentes, abhinc mox liheri, erecti, triquetro- 
acuminati, appendiculis 2 lin. longis. Stamina 5, in fauce inserta. Stylus apice 
bifidus. Ovarium biloculare ; placentae peltatge, multi-ovulatae. 
At the top of a ridge, alt. 6600 feet {Raviland, 1238). 

Very near to X. montana^ Korth., to ivhicli I believe a specimen collected by Motley 
(no. 659) near Banjermassin to belong, of which, however, I have not seen the type. 
It differs in the thicker branches with a whitish thin bark which early peels off 
irregularly, in the purplish acuminate stijiules, in the more coriaceous veinless leaves 
with very inconspicuous nerves, and probably also in the larger flowers. Prom 
L. morinda^ DC., it differs in the narrower and thicker leaves with very inconspicuous 
nervation, in the solitary heads supported by cup-shaped bracts, and in the long free tips 
of the corolla-lobes. 

Lucinji:a nehvulosa, Stapf, n. sp. Frutex scandens, glaberrimus. Eami novelli 
tetragoni, nigricantes, annotini cortice papyraceo albido, mox plus minusve soluto, 
laxe vestiti. Folia petiolo f poll, longo suffulta, ovato-oblonga vel elUptica, 3-3J 
poll, longa, 1^ poll, lata, basi rotundata, rarius brevissime in petiolum attenuata, 
breviter acuminata vel obtusa (ob apicem vulneratum ?), supra nigricantia, subtus 
olivaceo-fusca, nei^vis lateralibus x^rimariis distinctis idrinqtw circiter 11, sectmdariis 
tnterjectis smpe cBqualiter distinctis, venulis laxe et plus minusve conspicue 
reticulatis. Capitulum terminale, solitarium, pedunculo J poll, longo suffultum, 
bracteis binis obtusissimis in cupulam patelliformem connatis instructum, circa 12- 



DE. 0. STAFF OX THE FLOEA OF MOUXT KIJNABALU. 175 

florum ; flores arete congesti, sed liberi. Calyx ovoideo-cupularis, 3 liii. Ioniums ; 
limbus truncatus. Corolla) tubus 3^ lin. loiigus, su])erne dilatatus, intus ad :j alho- 
villosus ; lobi 5 lanceolati, 2J lin. lougi, crassi, appeudicibus in alabastro sjrpc liboris, 
brevibus, carinato-triquetris, carina basi in denticulum deilexuni productii. IStainina r> ; 
filamenta | lin. longa ; antliera? oblongo-lineares, 1^ lin. longa), exserta^. Stylus 
apice clavatus, bifidus. Ovarium biloculare ; placenta? peltatue, muiti-ovulatie. 

At 7000 feet (Ilamland, 1239). 

Closely allied to Liicincea pentacme and to L. montana, but diirering- from both in the 
larger leaves, from the former also in the very distinct venation, and from the latter in 
the loose papyraceous bark, the shorter jDcduncles, and the larger flowers. 

MTKiONErRo:N' BORNEEK^SE, Stapf, n. sp. llerba. Caulis cra^siusculus, tcjiuissime rufo- 
tomentellus, demum glabrescens. Folia aj)icem versus congesta, petiolo 1-2 jwll. 
longo sufFulta, lanceolato-oblonga, 7-10 poll. longa, 2-3 jjoll. lata, basi lonfi-e 
attenuata vel 2:>lus minusve in petiolum angustissime decurrentia, apice acuminat^i. 
membranacea, supra nigricantia, infra fusca, utrinque parce pilis tenuibus rufis 
adpressis conspersa, nereis lateral/bus utrluquc 8-10 e hasl Cffrrata suhrcclis, raUle 
p7'orsus cluclls, venis distinctis subtranscersalibus. Inflorescentia^ ex axillis superi- 
oribus pedunculo 3-4 lin. longo crassiusculo suffultae, capltatim coarctatcCt bracteatae. 
Bractese extimse foliacese, e basi latissima ovatae, tenuiter acuminata), 8 lin. longa), 
demum evanida), capitula involucrantes, sequcutes ovatae, interiores oblanccolata), 
flores brevissime pedicellatos a^quantes, rufo-pilosoe. Calyx rufo-pilosus ; tubus 
breviter turbinatus, IJ lin. longus; lobl 5, Unearl-Uuweolati, 2 lin. longi, acuti. 
Corolla alba ; tubus lobis calycinis aiquilongus, intus in fauce pilosus ; lobi 5, ovati, 
breves, f lin. longi. Stamina 5, sub fauce inserta ; filamenta tenuia, vix \ lin. 
longa; antherae lineari-oblongae, apice exsertae, | lin. longae. Fructus subbaccatus, 
didymus ; pyrenss binse crustaceoe, vix IJ lin. longte, teretes. Semina numerosa, 
angulata, minute granulata, placenta mediae axi affixa. 
At 3000 feet (JIavlland, 1332). 

It seems to be nearest allied to an undescribed species from the Admiralty Islands 
(' Challeno-er ' Exp.), whicb I am inclined to refer to tliis genus, and to another species 
from the Misbmi Hills {Griffith, 2850), also midescribed. The latter has, however, much 
looser inflorescences and narrower calyx-lobes, and approaches more If. niUans, Wall., 
whilst the former difl'ers in the venation of the leaves and is particularly distinguished 
by the laciniate bracts. 

Urophtllum slbakurum, Stapf, n. sp. 2 . Frutex. Eami obtuse tetragoni, badii vel 
superne exsiccando nigricantes, glaberrimi, vel noveUi sparse adpresseque pilosuli. 
Folia petiolo 4-8 lin. longo robusto sufFulta, omta vel elliptica, 3-3^ poll, longa, 
1^-2 poll, lata, basi breviter acuta vel rotimdata, apice breviter vel vix acuminata, 
subcoriacea, titrinque glabra vel subtus in costa valida et in nervis primo pilis fulvis 
adpressis induta, nervis lateralibus utrinque distinctis, subtus prominentibus, patulis, 
fere ad marginem leviter arcuatis, abhinc abrupte prorsus flexis, iitrinque circiter 10, 
3-4 lin inter se distantibus, venulis transversis, flexuosis, creberrimis, prominulis, 

2a2 



176 DE. 0. STAFF OX THE FLOEA OF MOUIS'T KINABALU. 

elongato-triangiilares, | poll, longse, hirtellse. InflorescentiaB pseiido-iimbellatse, 
4-5-floi'3e, pedimculo 5-1 lin. longo adpresse hirtello sufFiiltse; bracteae psendo- 
Terticillat8G, lanceolatse, 1-2 lin. longa?, liirtellse ; pedicelli J-J poll, longi, divergerites- 
Calyx liemisphferico-turbinatus, nigricans, glaber; tubus IJ lin. longns; limbus 
breviter 5-dentatus, dentibus late triangularibus. Corolla alba, late breviterque 
campanulato-tubulosa, ad f 5-lobata, deinde profundius fissa, 3 lin, Tonga, crassa, 
extus glaberrima, fauce villosa ; lobi triangulari-ovati, acuti. Stamina 5 ; filamenta 
toro cum corolla inserta, libera vel medio tubo plus minusve adnata, sed facile 
separata ; antherse oblongse, poUine destitutse ; connectivum dorso vix gibbosum in 
apiculum productum. Stylus crassiusculus, 2 lin. longus, lobis 5 linearibus sequi- 
longis. Ovarium 5-loculare; placentae peltatse, multi-ovulatse. Bacca lutea, 
siccando nigricans, globosa, limbo coronata, 3 lin. dimetiens. 
At 10,500 feet (Harilcmd, 1074). 

Uro]}'hyllum subanurum belongs to the group of IJ. corymbosum, Korth., which is repre- 
sented in Malaya by numerous species, most of which are still undescribed. It is 
distinguished chiefly by the firm, broad, and very shortly acuminate leaves, and by rather 
large flowers. These are, like those of the following species, dimorphic. The apparently 
normal anthers of the female are absolutely destitute of pollen when examined in bud. 

Urophyllttm CYPHANDurM, Stapf, n. sp. Frutex. Kami vix tetragoni, glabri vel 

primo adpresse hirtelli, interdum nigro-purpurascentes. Eolia petiolo ^ poll, longo, 

glabro vel hirtello suffulta, ovato-oblonga vel lanceolato-ohlonga, 4-6 poll, longa, 

1^-2:1 poll, lata, basi breviter acuta vel subrotundata, apice longiuscule caudato- 

acuminata, firme membranacea, plerumque plus minusve purpurascentia, imprimis 

in nervis, glabra vel subtus in costa valde prominula et in nervis hirtella, nervis 

lateraUbus utruique 9-10, 4-6 li7i. inter se distantibus, fere a basi longe prorsus 

arciiatis, prosilientibus, venis transversalibus flexuosis distinctis. Stipulye oblongse 

acutse, 6 lin. longa?, hirtellse. Mas : Inflorescentise pseudo-umbellata?, 6-9-florse, 

pedunculo hirtello 4-6 lin. longo suffultse, basi bracteis pseudo-verticillatis lineari- 

lanceolatis 2-3 lin. longis instructse; pedicelli 4-8 lin. longi. Calyx breviter 

cupularis, \\ lin. longus, nigricans, sparsissime hirtellus ; limbus 5-dentatus. 

Corolla alba breviter tubulosa, ad medium fere 5-loba, extus glabra, fauce villosa ; 

tubus 2 lin. longus ; lobi ovati, acuti, subsequilongi. Stamina 5 ; filamenta cum 

corolla toro inserta, paulo sub fauce cum tubo coh^erentia, cseterum libera; 

antherae poUiniferge, elliptic^e oblongse, f lin. longce; connectivum p>ostice valde 

gibbum, in apiculum productum. Stylus rudimentarius, brevis, laciniis 5 filiformibus. 

Ovarium minute 5-loculare, loculis vacuis vel placenta rudimentaria ovulisque 

paucis. Femina : Inflorescentia ut in mare, sed gracilior, glabra, 5-2-flora ; pedicelli 

demum ad | poll, longi. Calyx subglobosus, 1-1| lin. longus ; limbus basi constrictus, 

breviter 5-dentatus. Corolla (in alabastro tantum nota) maris corolla? simillima, sed 

minor. Stamina iis maris similia, multo minora, inania. Stylus crassiusculus, 

laciniis 5 lineari-lanceolatis. Ovarium 5-loculare ; placentae peltatae, multi-ovulatae. 

Bacca globosa, limbo calycino coronata, 21 iin. dimetiens. 



DE. 0. STAFF Oy THE FLOEA OF MOUNT KIXABALU. 177 

At 7800 feet [HavUand, 1123 $ , 1124 d }. 

Closely allied to the preceding species, Imt sufficiently distinct l)y the shape and 
nervation of the leaves, hy the much smaller female llowers, and by the s^ihbous shorter 
anthers. The sexual difference is Avell marked hy tlie different shape of the calvx, l)y 
the size of the flowers, and hy the development of the anthers and the stigmata. The 
insertion of the filaments is very curious ; but this mode is probably not so rare as one 
would suppose. Clarke has stated, in Journ. Linn. Soc. xvii. 159, that thev are free 
in the female flower of Adenosacme lougifolia^ and II. Baillon indicates a similar 
condition for Acrantliera zeylanica and Mnssccnda Landiw. 

UnopnYLLrM LiNEATUM, Stapf, n. sp. Erutex liumilis. llami obtuse tctragoni, robusti, 
adpressissime et densissime griseo-sericeo-velutini. Folia petiolo crassiusculo 2-3 
lin. longo suffulta, ovato-oblonga vel oblonga, 2-2^ poll, longa, 1-1} poll, lata., basi 
breviter acuta vel subrotundata, apice acuta, in margine (exsiccando ?) plerumque 
revoluta, crassa, coriacea, olivaceo-viridia, sujira costa pilosulii excejitii glaberrima, 
subtus in costd nervisque argcnteo-scricea^ ccctcrum glabra, nerms lateralihus lG-10 
subrectis, angnlo 45° oriis, 1-1^ Ihi. inter se distantibus^ supra impressis, subtus ut costd 
valde prosilientlbus, venulis transversis creberrimis inconspicuis. Stipula? lineari- 
lanceolata?, 3 lin. longae, subsericeoe. Inflorescentia^ pseudo-umbellatuL', sessilcs, basi 
bracteis ovatis vel lanceolatis 1-1 J lin. longis subsericeis instructa? ; pedicelli 1-1| lin. 
longi. Calyx nigricans, glaber vel subglaber, in ilore maseulo breviter cupulari- 
turbinatus, in flore femineo subglobosus ; limbus 4-, raro 5-dentatus, dentibus 
minutis. Corolla ignota. Bacca nigra, globosa, 4-5-locularis, limbo calycino 
coronata, 2^ lin. dimetiens. Semina numerosa, in placentis peltatis carnosis, 
distincte granulata. 
Maripari Spur, alt. 5500 feet (Haviland, 1266). 
A very fine plant, distinguished by the numerous closely set nerves, which are, like the 

midrib, very prominent and white silky. Though the nervation appears at the first 

glance very different from that of the described Urophyllums, it does not differ typically. 

The calyx, the berries, and the seeds are those of typical Vi^ophyllum. 

Ueophyllum SALiciFOLirM, Stapf, n. sp. Frutex. Rami graciles, teretes, dense 
adpresseque fulvo-griseo-pilosi. Folia pedunculo 3-4 lin. longo suffulta, lanceolata, 
21-31 poll, longa, 7-10 lin. lata, basi breviter acuta, apice acuminata, pergaraacea, 
supra obscure olivaceo-viridia, praeter costam pilosulam glabra, supra flavido-viridia, 
in costa et in nervis adpresse pilosa, cgeterum glabra, ncrvis lateralibus utrinque 
12-13, 2-3 lin. infer se distantibus, prominentibus, panlo supra basin valde prorsus 
iiexis et usque ad nermim collectivum valde distitictum rectis, venis transversis 
creberrimis, tenuissimis. Stipulse lineares, 3-4 lin. longse, dense pilosae. Inflores- 
centi^e ad Jlores 1-2 basi bracteatos reductse, supra-axillares ; pedicelli 1 lin. longi, 
adpresse pilosi. Calyx pilosulus ; tubus globosus, vix 1 lin. longus ; limbus lasi 
constrictiis ad f 4-dentatus, dentibus late triangularibus, f lin. longis. Ovarium 
4-loculare ; placenta peltatse, multi-ovulatse. Bacca globosa, 2 lin. dimetiens 
dentibus ob margines involutos subulatis coronata. 



178 DE. 0. STAFF ON THE PLOEA OF MOUNT KINABALU. 

At 6000 feet (Ilaviland, 1231). 

I have described this plant notwithstanding the imperfect material, as there is no doubt 
about its generic affinity, and because I think it is sufficiently marked by the 
characteristic nervation, the very reduced inflorescence, and the comparatively large 
membcanaceous caly-xlimb. Apart from the inflorescence it resembles UrojjJii/U 
streptopodium, Wall., but the nerves are more numerous, closer, and straighter. 



um 



TJROPHTLLrM LONGiDEKS, Stapf, n. sp. $ . Frutex. Rami teretes, densissime adpresse- 
que flavido-cano-hirtelli. Polia petiolo eodem indumento vestito 5-7 lin. longo 
suffulta, lanceolata vel oblongo-lanceolata, 4-4|^ poll, longa, 1 poll, lata, basi acuta, 
apice acuminata, subcoriacea, flavido- vel cano-viridia, supra costa nervisque adpresse 
hirtellis exceptis glabra, opaca, subtus in nervis venisque adpresse strigi^loso- 
hirtella, nervis lateralibus utrinque 11-12, angulo 40° ortis, subrectis, sub margine 
arcuatim connexis, nervo circumcirca marginante addito, venis transversis flexuosis, 
supra tenerrimis, subtus validioribus, prominentibus. Stipulse e basi triangulari 
lineari-subulata?, 6-7 lin. longa3, hirtellse, mox deciduaB. Cymulse S-l-florae, 
sessiles ; bracteae triangulares, 1 lin. longse; pedicelli hirtelli, 2-3 lin. longi. Calyx 
adpresse hirtel'us; tuJms ovoideus vel obovoideuSy 1-^ lin. longtis; lobi 4 linear es, 
\-\\ I'm. longi. Corolla viridi-alba; tubus IJ-lg lin. longus, pro ratione latus, 
extus glaber, in fauce villosus ; lobi 1, ovati, tubo paulo breviores, extus apice pilosiili. 
Stamina imperfecta 4 vel pauciora, fauci inserta ; filamanta ^ Jin. longa ; antherae 
minute cordatse, effetse. Stylus ciassiusculus, superne dilatatus, 1^ Jin. longus ; 
stigmatis lobi 4, lineari-oblongi, |-1 Jin. longi. Ovarium 4-locr^are; ovu^a numerosa 
in placenta media peltatini affixa. 

At 7600 feet {Raviland, 1120). 

One of the best marked species the affiuHy of which I find difficult to point out. It 
comes probably nearest to U. scdicifolium^ Stapf. 

Randia dexsifloea, Benth., Fl. Hongk. 155. 

A small tree, near Korng, at 2000 feet [Ilaviland, 1380). 

Distribution: >U over Malaya, Indo-China, to the Philippines and North Australia; 
on the mainland from Singapore to Cachar and Upper Assam and in Travancore ; South 
China. 

IxoEA KINABALU ENSis, Stapf, n. sp. Frutcx, inflorescentiis exceptis, glaberrimus. 
Eamuli graciles, pallidi. Folia pedunculo J-f poll, longo suffulta, oblongo-lanceo- 
lata, 4-5 poll, longa, lh~2 poll, lata, basi acuta, apice acuminata, coriacea, supra 
lucida, exsiccata olivaceo-viridia, nervis lateralibus utrinque 10-12, angulo 50-60° 
ortis, leviter arcuatis, nervo collectivo distincto, supra impressis, infra prominulis. 
Stipulse e basi breviter lateque ovata abrupte subulatse, sine subula 1 lin. lon"-8e. 
Inflorescentia terminalis, brachiato-corymbosa, pendula, peduncido 4-7 poll, 
longo, minute puberulo suffulta, supra basin et ad ramos infimos bracteata, 
puberula. Bractese lanceolatse 1-2^ lin. longse, inferiores more foliorum stipulatae. 
Pedicelli 1-2 lin. longi, graciles, basi pleruwqtie nudi, vel rarissime minute bracteo- 



DE. 0. STAPP OX THE PLOliA OF MOUXT Klx\A]UT.U. 179 

lati. Calyx ovoideus, vix 1 liu. longus, glaber, dontibiis i brcvibus rotimdatis 
Corollae tubus ruber, gracilis, 9 lin. longus ; Johi fore all)!, oboralo-ohlongi, 3 lin. loiii^i, 
obtusissimi. Filamenta latiuscula, J-j liu. longra ; anthortc sagitbita% 2 liu. loni^w. 
Stylus circiter 1^ lin. coroUam superans ; stigma bilobum, lobis crassiusculis. Bacca 
globosa vel subdidyma, 4J lin. dimetiens, rel transverse latior, loculo altero s£epe 
abortivo. 
At 5000 feet {narilmid, 1236). 

Closely allied to Lvo)Yt pendida, Jack, but distinct by the smaller leaves and chiefly by 
the always distinct pedicels, which very rarely are supported by minute bracteoles, and 
by the shorter corolla with broader lobes. In I. pendula the flowers are constantly 
sessile and supported by one or (if terminal) by two minute bracteoles. 

Pavetta loibata, Stapf, n. sp. Prutcx humilis. llami novel! i siibteretes, nigrescentcs, 
puberuli, mox glabrati, annotini pallidi, internodiis pleriscpic brevi])us, 2-5 lin. 
longis. Folia petiolo 2-3 lin. longo suffulta, lanceolata, 3-3^ poll. longa, ^-1 poll, 
lata, basi cuneatim attenuata, apice acute acuminata, membranacea, glabra, luride 
olivacea, nervis lateralibus utrinque circa 7, 5-9 lin. inter se distantilms, tenuibus, an- 
gulo 45°-40°ortis, in nervum subniarginalem collectis. Stipuhe late ovatie, mucronataB, 
2 lin. longee. Inflorescentia terminalis, pubescens, tliyrsoideo-paniculata, juvenilis 
1-1 J poll, longa, bracteata ; bractece ad 1 lin. long{c, subulata^ ; flores pedicellati; 
pedicelli lJ-2 lin. longi, graciles. Calyx rainutissime puberulus, tubus ovoideus, 
■| lin. longus ; limbus ceqidlongus, suhmembranaceiiSy e basi coiistrictd cupuU form iter 
ampUatus, ^-lobnlafus, lobuUs e basi constrictd latissimc ovatis, (ramrerse latioribus, 
minute ciliolatis. Corollae (ex alabastris 2-3 lin. longis fere perfectis tantum notai) 
tubus brevis ; lobi 5, lineari-ol)longi. Stamina 5 ; filamenta brevia ; antheras lineares, 
vix 2 lin. longse. Stylus filiformis, superne incrassatus, simplex. Ovarium biloculare; 
ovula solitaria medio afiixa. 

At 6000 feet (Haviland, 1240). 

Allied to P. indica, L., but the calyx is very different in the comparatively wide 
submembranaceous limb and the very short but broad minutely ciliatc lobes. 

Pstchotria malatana, Jack, in Mai. Misc. i. 3. 

Low land, on the Tawaran River {Haviland^ 1389). 

Distribution : Malaya : Penang, Sumatra (?), Java, Borneo. 

I consider this plant as identical with Jack's P. mahnjana, as it is represented by 
"Wallich's P. atirantiaca, no. 8335 (coll. by W. Jack, 1819, at Penang), and it is certainly 
identical with various specimens from North Borneo, as Motley's, Labuan (218), and 
Dr. M. Eraser's, Kudat (86). The lower cyme branches are often opposite, but sometimes, 
like the upper ones, wborled. I refer also to it Zollinger's specimens (173) ; a specimen, 
sent by Teysmann from Java, named Chasalia expausa, Miq. ; and one from Miquel's her- 
barium named P. auranfiaca, Miq. A capsule with loose fruits attached to Teysmann's 
specimen contains two kinds of fruits. Two of them, which are more elongate, have 8 deep 
furrows and do not belong to this species, as Sir J.Hooker has suggested. The remaining 



180 DK. O. STAPr O^ THE PLORA OF MOUNT KINABALU, 

fruits agree perfectly witli those of the Bornean plant, and, as a comparison with the 
young fruit still attached to the hranch proves, they belong doubtless to Teysmann's 
specimen. But the true Ch. expansay Miq., or Psychotria expansa, BL, is a different 
plant with membranaceous leaves and identical with P. montana, Bl. The fruit is 
slightly compressed from the back and hardly furrowed at all. The endosperm is very 
ruminate. A specimen collected by Motley at Banjermassin (608), and probably 
identical with P. crassifolia, Miq., comes also very near, but it has smaller and more 
coriaceous leaves. The inflorescence and the fruits, however, are exactly as in P. 
malayana. 

The Kinabalu specimens represent the microstyle form. The stigmata, which are a 
little narrow^er and perhaps less papillose than in the macrostyle form, almost reach 
the mouth of the corolla, whilst the filaments surpass the mouth to the extent of J-1 lin. 
The anthers are about 1 lin. long. This form is represented also by a part of Zollinger's 
specimens (173), whilst Teysmann's and those of Eraser and Motley belong to the 
macrostyle form. Here the style surpasses the mouth by more than 1 lin., whilst 
the anthers, which are also 1 lin. long, just reach the mouth by their tips. The 
filaments are extremely short. The microstyle flowers were observed in inflorescences 
with half -mature berries. The anthers, which in both forms hardly differ, were always 
found to bear apparently quite normal pollen. A curious intermediate state is re- 
presented by another of Zollinger's specimens, also numbered 173. It is evidently 
macrostyle ; but two of the stamens are exserted, 3 included. They, however, do not 
seem to be quite normal. The stigmata are long, linear, and minutely papillose. The 
flowers exhibit another abnormality. The corolla is 4-lobed, apparently by the 
coalescence of two lobes, whilst the androecium is pentamerous. The exserted stamens 
stand one on each side of the double lobe, which is hardly larger than the other lobes, 
whilst a small stamen is opposite to it. The remainder of Zollinger's specimens is 
pentamerous, like all other specimens I have seen, except those of Eraser's and Motley's 
collection, W'hich are tetramerous. 

PsTCHOTniA GYRTiLOSA., Stapf, n. sp. (§ Gvumilea). Frutex scandens(?). Bami glabri, 
demum pallidi. Polia petiolo 4-5 lin. longo suffulta, lanceolata, 3^-7 poll, longa, 
1-lJ poll, lata, basi cuneatim attenuata, acuta vel subacuminata, pergamacea, 
glabra, supra nigricantia, subtus nitore rubello suffusa, nervis lateral! bus 9-12, 6-8 lin. 
inter se distantibus, arcuatis, valde prorsus ductis, marginem versus venis transversis 
tenuibus. Inflorescentia terminalis, paniculata, pedmiculo 2| poll, lon^-o stricto 
suffulta, ramis verticillatis quaternis, cymulas pancifloras suhumbellatim cUspositas 
gerentibus. Bacc^e sesslles, pedimculis ad 2 lin. lougls, pedicellos imitantibus 
insidentes, globoso-pyreniformes, 3-4 lin. Ion gee, indistincte ruguloso-costata?, calycis 
limbo parvo coronatse. Vyrence crustacese, intus gyroso-rugulosae, rugis l^vissimis 
prsesertim inter costas transverse directis et varie inter se confluentibus. Albumen 
rugis conformatum, gyroso-ruminatum. 

At the Penokok, alt. 3200 feet (Baviland, 1335). 

There was a flowering branch sent on the same sheet and with the same label. It is 



DE. O. STAFF OX THE FLOEA OF 310UNT KIXAB.ALU. ISl 

very likely that it belongs to the same species as the fruiting branch, but as there are 
some differences in the leaA'es and in tlie inflorescence I do not feel quite sure of it. 
The following is a description of it : — 

Folia petiolo 5-6 hn. longo sufTulta, inferiora lauceolata, 2A-1J poll. longa, f- 1^ poll, 
lata, basi acuta, apice subacumiuata, nervis lateralibus utrinque 7-11, 4-5 lin. inter 
se distantibus, superiora multo majora, oblonga vel obverse lanceolato-oblono-a, 6-8 
poll, longa, 2J-2f poll, lata, basi acuta, apice obtuse acuminata, nervis 15-16, 4-6 lin. 
inter se distantibus, augulo circa SO^-eO" ortis, subrectis, sub margine breviter 
arcuatim in nervum collectivuni connexis, venis tenuibus transversis laxis, omnia 
glabra, supra obscure olivacea, subtus pallidiora, nitore rubello inconspicu ). Stipube 
ovatae, ad f connatas, obtusissimte, vix 4 lin. longae. Inflorescentia terminalis, panicu- 
lata, minutissime puberula; pedunculus 5 lin. longus ; rami oppositi, summi verticil- 
lati, infimi 1 poll, longi; cymuUv pauc'ijlorcc suhumbellalim dispositw ; fiores sessilea 
vel i^seudopedicellati^ alares hracteolis minutissiniis 2 instructi. Calyx cupultiri- 
turbinatus, 1 lin. longus, glabcr; limbus truncatus vel minutissime denticulatus. 
Corolla; tubus 2 lin, longus, rectus, fauce villosus, viHo vix exserto ; lobi 4, ovato- 
oblongi, 1;^ lin. longi, patuli vel reflexi. Stamina 4 exserta ; filamenta 1-1 J I'm. 
faucem superantia ; anthercc | lin. longer, oblongo-lineares. Stylus inclusus ; 
stigmata oblonga (semper?) coluL'rentia faucem vix attingentia. 
This species is certainly allied nearly to P. aunintiacay but distinctly different in the 
looser inflorescence, and the smaller flowers with much smaller more exscrted antiiers^ 
and smaller berries. The flowers are all sessile, but, as the lateral flowers of the outer- 
most cymes are suppressed, the central flower is borne by the peduncle of the reduced 
cyme and supported by the minute bracts of the dichasium. 

PsYCHOTRiA ITEOPHYLLA, Stapf, n. sp. Frutcx humilis, glaberrimus. Rami teretes, 
linea e conimissura decurrente j)i*ominente notati, nigricautes. Folia petiolo ^-f 
lin. longo suffulta, lanceolata, 4-5 poll, longa, f-1^ poll, lata, basi longe cuneatim 
attenuata ; apice sensim acuminata, obscure undulata, supra obscure olivacea, 
subtus rubro-suffusa, nervis lateralibus utrinque 9-10, patulo-erectis, leviter 
arcuatis, in nervum marginalem plus minusve distinctum coUectis, laxe et imperfecte 
reticulata. Stipulae e basi ovata longiuscule acuminata?, 3-5 lin. longa3. Inflores- 
centia terminalis, corymboso-paniculata, IJ poll, longa, 2 poll, lata; pedunculus 1 ])oll. 
longus ; rami oppositi, inferiores iteratim di- vel trichotomi ; cymuUe paucijlorce, 
laxce, minutissime bracteatae ; pedicelU graclles, 1-1\ lin. longi. Calyx cupulari- 
turbinatus, f lin. longus. Corollas albae tubus J lin. longus, pro ratione latus, in 
fauce lined mllosd notatus ; lobi 5, ovati, 1 lin. longi, reflexi. Stamina 5 ; filamenta ad 
basin corollas distincta et facile separata (an libera?), circiiev f lin. faucem super- 
antia ; antherae ellipticae, vix \ lin. longffi. Stylus 1 lin. exsertus ; stigma clavato- 
incrassatum. Bacca rubra, ellipsoideo-globosa, 3 lin. longa, 8-sulcata; pjyrence 
tenuiter crustacece, dorso profunde sulcatae. Albumen aequale, endocarpio confor- 
matum, sulcatum. 
At the Penokok, alt. 3500 feet {Haviland, 1321). 

SECOND SERIES. — BOTANY, VOL. IV. 2 T, 



182 DE. 0. STAPE ON THE FLOEA OF MOUNT KINABALU. 

Allied to Fsychotria erratica. Hook, fil., and to P. mridiflora, Eeinw., but differing from 
the former in the generally narrower leaves, the looser inflorescences, and the pedicelled 
flowers witli more exserted anthers and a much less yillous throat, and from the latter, 
\\liich it approaches indeed very closely, in the narrower leaves, but chiefly in the 
larger, more oblong berries, with a much thinner endocarp. 

PsYCHOTEiA DENSIFOLIA, Stapf, n. sj). (PL XIII. B. 11-13.) Frutex humilis, plerumque 
epiphyticus, glaberrimus. Bami teretes vel obscure tetragoni, novelli obscuri, 
annotini cortice sulcato cinerascente vestiti, internodiis 4-5 liii. longis. Polia petiolo 

1 lin. longo vel breviore snffulta, ovato-elUptica vel elliptica, 8-10 lin. longa, 6-7 
lin. lata, basi breviter acuta vel subrotundata, apice acuta vel obtusiuscula, coriacea, 
flavido-yiridia vel fuscescentia utrinque liicida, nervis lateralibus utrinque 5, angulo 
45° ortis, subrectis, tenuibus, subevenia. Stipulge e basi connata brevissime 
latissimeque ovatse, apiculatse, purpurascentes, deciduae. Inflorescentia terminalis, 
corymbulosa, peduncvilo 2-4 lin. longo snffulta, \ poll, longa, f poll, lata; flores 5-8 
umbellatim congesti in ramulis 2-4 lin longis ; pedicelli sub anthesi 1 lin. longi. 
Calyx cupulari-turbinatus, 1 lin. longus ; limbus 5-dentatus. Corolla viridula, 
extus minute papillosa ; tubus latiusculus, superne sensim dilatatus, 1 lin. longus ; 
lobi 5, ovato-oblongi, demum reflexi, f lin. longi. Stamina 5 ; filamenta ad medium 
tubum corollas adnata, J lin. exserta ; antherae oblongse, vix J lin. longae, albidae. 
Stylus inclusus, cum stigmate clavato f lin. longus. Bacca ellipsoideo-globosa, 

2 J lin. longa, calyce persistente coronata, indistincte costata ; pyrenae tenuiter 
crustacese, distinctius 4-sulcat8e. Albumen (seminis semimaturi) indistincte 
ruminatum. 

At 8000 feet {miviland, 1116). 

Nearest aUied to F. polycarpa, Hook, fil., but one of the best marked species, distin- 
guished by the crowded, small, and glossy leaves, and the short inflorescences. 

Stkeblosa tjrticina, Stapf, n, sp. (PL XIII. a. 1-10.) Herba scandens. Caulis teres, 

prime minutissime furfuraceus, mox glabratus, olivaceo-fuscus, internodiis lineis e 

commissuris decurreutibus prominulis notatis. Eolia petiolo J-1 J poll, longo snffulta, 

oblongo-lanceolata, 5|~7i poll, longa, 2-2J poll, lata, basi longe attenuata, sub- 

acuminata, tenuiter membranacea, exsiccando olivaceo-raembranacea, subtus in 

nervis minutissime furfuraceo-puberula, caeterum glabra, nervis lateralibus utrinque 

12-15, erectis, inferioribus approximatis. Stipulae ad basin bifidae, segmentis e 

basi triangular! subulatis 1 lin. longis. Paniculae axillares petioles a3quantes, 

pedunculo \ poU. longo suffultae, dense minu.teque furfuraceo-hirtellae ; rami simplices 

vel dichotomi, graciles, interinipte suhspiciformes ; Jlores in ramis vel ramulis 

sohtarii vel geminati ; hractecB suhtendentes mimitce, alice oppositce pmnlo majores ; 

pedicelli brevissimi. Calycis limbus truncatus, minutissime 4-denticulatus, minute 

hirtellus, persistens, J lin. longus. Corolla (ex alabastris tantum nota) sestivatione 

valvata vel, ob loborum margines exteros interduin perpaulo ampliatos, specie sub- 

imbricata ; tubus brevis, intus basin versus pilosulus. Eilamenta glabra ; antherte 

oblongaB. Discus depresso-globosus, raagnus. Ovarium biloculare ; stylus glaber ; 



DE. O. STAFF ON THE FLORA OF MOUNT KINABAI.U. 183 

stigma bilobum, lobis ovatis; oviilci solitaria, fere ex iina basi erecta. Fructus 

globosus, 1 lui. dhjietiens, jugis 8 proiiiiuulis, minutissiine hirtellus, (lii)ynMuis ; 

pyrena altera minor vacua ; endocarpium crustaceum iu facie veritrali sulcatum vel 

plus minusve bisulcatum. Albumen aiquale. 
At the Penokok, alt. 3200 feet [Ilaviland, 1328). 

I think Streblosa should be kept up as a distinct genus on account of its peculiar 
habit, which is produced chiefly by the herbaceous growth, by the very tliiii leaves, the 
axillary inflorescences (which in the present species resemble very mucli those of a 
common nettle), and by the minute flowers and fruits. Miquel snys of Streblosa^ 
"frutices"; but the specimens I saw of JPsycJiotria tort His, Bhiiuc, the only species 
seen by him, are decidedly herbs. The genus Streblosa is limited to ]Malaya. 

G^RTNERA RUFiNERVis, Stapf, n. sp. Frutcx. Ramuli gracilcs, novelli sparse rufo- 
pilosuli, demum glabrati, cortice cinerascente. Folia petiolo plcrumque 2-1 lin., 
raro ad 7 lin. longo suffulta, lanceolata vel oblanceolata, 3-4 poll, longa, 1 poll, lata, 
longe in petiolum attenuata, apice acuminata, tenuiter coriaceii, exsiccando rufo- 
fusca, supra glaberriraa, subtus primo in nervis plus minusve rufo-furfuracca, mox, 
axillis nervorum hirsutulis excei)tis, glabrata, nervis lateralibus utriucpu' 7-1) supra 
leviter impressis, subtus prominentibus, sub margine arcuatim collectis. Stipul;e 
2-3 lin. longse ad medium tubulose commtcey demum profundins fissa% apice late 
rotimdatce, ciliatae. Panicula pedunculo J-f- poll, longo suffulta vel subsessilis, 
brachiata, 1-1 j poll, longa lataque, f urfuraceo-puberula ; cymula} denstc in corymbum 
dispositae ; bracteae minuta3, acuta? ; pedicelli brevissimi. Calyx turbinatus, paulo 
ultra J lin. longus; lobi j^ofiindati, minuti, ciliatuli. Corolla? tubus 1 lin. longtis, iu 
fauce dense villosus ; lobi late ovati, vix f lin. longi. Stamina breviter exserta, 
J lin. longa ; iilamenta antheris paulo breviora. Stigmata crassiuscula. 
At 6000 feet [Razikmd, 1210). 

Allied to G. Jimghuhniana, Miq., but chiefly distinct by the shape of the stipules, by 
weaker panicles, smaller flowers (about half the size), and by the shape of the calyx- 
limb. 

Chasalia gracilis, Stapf, n. sp. Frutex glaberrimus. Rami gracilcs, teretes, lineis e 
commissuris decurrentibus tenuibus notati, viridi-albidi. Folia petiolo tenui, 
2 lin. lonsco sutfulta, ovato-lanceolata, 2 poll, longa^ 7-0 lin. lata, basi citncata, apice 
acuminata, tenuiter membranacea, in margine circumcirca tenuiter incrassata, 
pallide olivacea, nervis lateralibus utrinque 7-9, tenuibus, inferiorilms 2-3 lin. inter 
se distantibus. Stipulae e basi connata brevissime ovatae, pallida?, vix 1 lin. 
attingentes. Inflorescentia terminalis, corymbulosa, pedunculo 1-1 J] poll, longo 
o-racili suffulta, minute bracteata ; bractea? subulatse ; flores pedicellati ; pedicelli 
i_^ lin. lono>i. Calvx cupulari-turbinatus, f lin. longus ; limbus 5-denticulatus, 
minutissime papillosus, Corollse tubus fauce villosus, pro ratio ue latus, | lin. 
lono-us ; lobi 5, oblongi, sequilongi. Stamina 5, sub fauce inserta ; filamenta 
brevissima ; antherse ellipticse, \ lin. longse. Stylus e fauce breviter exsertus, 1 lin. 

2b2 



184 DE. O. STAFF ON THE FLORA OF MOUNT KINABALU. 

longus ; stigma bilobum, lobis latis obtusis. Bacca (immatura) subglobosa, 2 lin. 
longa, calycis limbo coronata, minute albo-lineolata. 
At tbe Penokok, alt. 3500 feet [Baviland, 1294). 

Closely allied to Chasalia rosti^ata, Miq., but the leaves are thinner and of a different 
shape, and smaller than in any specimen of C. rostrata I have seen. 

Lasiakthus kixabaltjensis, Stapf, n. sp. Prutex glaherrinms. Rami graciles, teretes, 
flavido- vel fusco-virides. Polia pedunculo 2-3 lin. longo suffulta, elliptico-oblonga 
vel oblongo-lanceolata, 3-3J poll, longa, f-lj poll, lata, basi acuta, apice subcaudato- 
acuminata, firmule membranacea, pallide viridia, nervis lateralibus utrinque 5-6, 
valde arcuatis, venis laxissime anastomosantibus, pcmcis, rectis vel snbrectis. 
Inflorescentise ad receptacula parva nodiformia sessilia redactse, pauciflorEe, axillares ; 
pedicelli nulli. Calycis tubus ovoideus vel oblongus, 1 lin. longus ; lobi 5-3, 
lineari-oblongi obtusiuscuU, sequilongi. Corollse tubus tenuis, IJ lin. longus, fauce 
villosulus; lobi 5-3, lineari-oblongi vel oblongi, inxus papillosi, tubo a?quilongi. 
Stamina 5-3, fauci inserta ; antherse subsessiles, oblongse, J lin. longse. Stylus 
exsertus, 2 lin. longus ; stigma bilobum. Ovarium 5-3-loculare ; ovula solitaria, 
basalia. Bacca globoso-pyreniformis, calyce coronata, 2 lin. longa ; pyrena unica; 
loculi cseteri abortivi (an semper ?). 
At the Penokok, alt. 3000-3200 feet (Havilcmd, 1306, 1323). 

Allied to L. lucidus, Bl., which it resembles very much in every respect ; but JL. hina- 
baluensis differs in the entire absence of all indumentum, in a somewhat different nervation 
— the veins being less straight and more loosely anastomosing, — and in the smaller 
flowers with obtuse calyx-lobes. The leaves are very minutely pellucid-dotted in 
consequence of the very loose spongy mesophyll. 

Lasianthus membiia:nacetjs, Stapf, n. sp. Prutex fflaberrimus. Bami teretes, pallide 
olivacei. Polia petiolo 4-7 lin. longo suffulta, oblonga vel elliptico-oblonga, 
5-7i i^o//. longa, 2J-3 jwll. lata, basi acuta, abrupte breviterque acuminata, 
membranacea, ^viX^\^Q lurideque viridia, nervis primariis utrinque 5-6 basi plerumque 
patulis, tunc valde prorsus arcuatis, secmidar lis tenuioribns plerumque 3-4 interjectis, 
cum venis laxissime reticulantibus. Stipule breviter lateque triangulares. 
Inflorescentise axillares ad receptacula nodiformia redactse ; alabastra plura, flores 
perfecti vero pauci ; pedicelli nulli. Calycis tubus obovoideus, f lin. lono-us, 
dentibus 4 triangularibus, tubo dimidio fere minoribus, crassiusculis. Corollte tubus 
IJ lin. longus, fauce villosus ; lobi lineari-oblongi 4, intus papillosi. Stamina 4 ; 
antherse subsessiles, lineari-oblongye, ^ lin. longse, inclusse. Stylus 6 lin. lon"-us, 
exsertus. Ovarium 4-loculare ; ovula solitaria, basalia. Bacca 4^7 lin. Ion ""a 
bipyrena ; pyrense insequales. 
At the Penokok, alt. 3000 feet {Havilmid, 1300, 1336). 
Allied to L. oUganthus, Hook, fil., from Ceylon, but perfectly glabrous, and the leaves 

more membranaceous, larger, and with a somewhat different venation. They are full of 

.Tiiiuite pellucid dots, as in the preceding species. 



DR. O. STAFF ON THE FLORA OF MOUNT KINABALU. Ib5 

Lasianthus MEMBEAis^ACEUs ?, Stapf, var. FiRMA, 11. var. Folia niinora elliptica, 3-.1 
poll, longa, l|-2 poll, lata, abrupte acuminata, firmiom, ncrvis ])lorunique utrinque 4 
minus patulis. Bacca ellipsoidea, utrinque breviter attouuata; pyreiuu crasse 
crustacese 2, quarum una minore vacua. 
On the Kadamaian, alt. 5000 feet (Ilaviland, 1241). 

In spite of the smaller and much firmer leaves, I believe this form should be referred 
to the species just described. The venation of the leaves is evidently of the s<*ime type, 
though the nerves are less spreading and fewer in number in accordance with the smaller 
size of the blade. Otherwise there is no difference, except in the firmer consistency of the 
leaves. These differences may be due to the higher elevation and a more exposed 
habitat. 

Lasianthi^s EUNEURrs, Stapf, n. sp. Frutex. Kami teretes, dense minuteque fulvo- 
tomentelli. Folia petiolo 3 lin. longo suffulta, oblonga, 3-3^ poll, longa, 1:^-1} poll, 
lata, basi breviter acuta et sa^pe leviter asymmetrica, apice acuta^ pergamacea^ 
supra nigro-fuscescentia, glaberrima, subtus in nervis minute densequc fulvo- 
tomentella, cieterum glaberrima, utrinque lucida, ncrvis lateralibus utrinque 7-8, 
longe prorsus arcuatis, in nervum marginantcm collectis, sccundariis paiicis brevibus 
interjectis cum venis transversis numerosis atque venidis ntruiqiie gracillime excisis 
pulchre reticulantibus. Stipulee e basi triangulari subulato-acuminatoe, tomentellai. 
InflorescentisB axillares ad flores paucos fasciculatos redacts, sessiles vel brevissime 
pedicellatse. Calyx breviter turbinatus, vix 1 lin. longus, tomentellus ; limbus 
minute lateque 4-dentatus, deciduus. Corolla ignota. Bacca globosa, vix 2 lin. 
dimetiens, 4-sulcata, (exsiccata) nigra, hirtella, 4-pyrena. 
At 7600 feet {Kamland, 1119). 

Nearly allied to L. acuminatus, Wight, and L, strigillosiis, Hook, fil., from tlie Western 
Peninsula, and perhaps also to L. Wightianus, Hook, fil., from Malacca ; but the leaves 
are firmer than in the first two species and not really acuminate, and the reticulation 
is more prominent and exceedingly sharp. L. Vy'ightiauus has more numerous 
nerves and a distinct reticulation. Amongst the Malayan species Z. stoxot^arius, 
Blume, comes nearest ; L. retioulatus, Blume, which evidently is also nearly allied, I do 
not know, except from the description, but it is said to have rather large acuminate leaves. 

Lasianthus rottjndatus, Stapf, n. sp. Frutex subglaber. Hami nigricantes, glaberrimi, 
tevetes, inte7vwdiis vix ^ poll, longis. Folia petiolo 1-1 -J poll, longo suffulta, late 
ovata, 1^-2 poll, longa, 1-1^ poll, lata, basi late rotundata vel suhcordata, apice 
acuta vel apiculata, coriacea, utrinque lucida, nigro-fuscescentia, subtus in nervis et 
in margine incrassato sparse hirtella vel glabra, nervis lateralibus utrinque 5, 
arcuatis, in nervum marginantem collectis, venis transversis numerosis crassiusculis 
venulisque reticulantibus utrinque prominulis. Stipulae triangulares, breves. 
Inflorescentiai valde ahbrevlatce, 1-2 lin. longa? (floribus exclusis), ssepe nodiformes, 
subhirtellse ; alabastra pro parte mox decidua vel abortiva, cicatrices relinquentia 
numerosas ; flores perfecti pauci, sessiles vel brevissime pedicellati. Calyx cupulari- 



186 DE. O. STAPP ON THE PLOliA OF MOUN^T KIXABALU. 

turbinatus, If lin. longus, sparse pilis minutis aspersus; limbus latissime brevis- 
simeque 4i-de)itatus, ciliolatus. Corolla parce pilosula ; tubus latiusculus, If lin. 
longus, fauce villosulus ; lobi 4, oyati, 1 lin. longi. Stamina 4, fauci inserta, exserta; 
filamenta nigra, crassiuscula, f lin. longa ; antherse oblongo-elliptica3, ^ lin. longse. 
Stylus longitudine tubi corollini ; stigma 3-4-lobum. Ovarium 4-loculare. Bacoa 
subglobosa, 4-loba, 2 J lin. longa, calycis limbo dentibus plerumque delapsis coronata ; 
pyrense 4, Crustacea?, secundum angulum internum excavatse. 

At 8300 feet {mivikmd, 1141). 

Nearest allied to Lasicmthus vemilosus, "Wight, but distinct by the rounded or even 
subcordate base of the crowded leaves, shorter inflorescences, and very short calyx-teeth. 
On the other side it is allied also to X. eimeurus, which has very similar fruits. 

Nebteea depeessa, Banks, in Gaertn. Pruct. i. 124, t. 26. 

Creeping in jungle at 10,500 leet {Saviland, 1075). 

Distribution : South and Central America, from the Falkland Islands and Patagonia 
to South Mexico ; islands of the South Sea to Tasmania and Victoria ; New Zealand ; 
Hawaii ; Malayan Archipelago to Sumatra and the Philippines. 

I think the Philippine plant ( Cuming, 943), as well as the specimens from Java (Mt. 
Malabar, 7000 feet, Anderson, 193 ; Horsjield, 779 ; Merapi, 3700-4000 feet, Junghuhn), 
must be referred to N. depressa, though they differ at the first glance more or less — but 
not more than do the specimens from New Zealand or those from Tristan d'Aeunha. 

CoPROSMA cRAssiCAL'Lis, Stapf, u. sp., 6 2 . Prutcx dioecus, prostratus, 4 ped. altus, 

glaherrimus. Hami obtuse tetragoni, novelli nigricantes, crassiusculi, annotini 

badii vel fusci, ad J poll, crassi. Eolia subsessilia, oblonga vel obverse lanceolata, 

f-l| poll, longa, 4-5 lin. lata, basi breviter vel longius attenuata, apice acuta vel 

apiculata, coriacea, nigricantia, subtus pallidiora et sublucida, in marginc anguste 

arcteque revoluta, nervis lateralibus utrinque circiter 5-6, tenuibus, supra impressis, 

subtus prominulis, vel utrinque indistinctis, venis nullis. Stipulse ovatcB, \\ lin. 

longse, obtusae, crasse fimbriatse, fimbriis minute ciliatis, albidse. Plores singuli, ad 

apices ramulorum interdum valde abbreviatorum subsessiles. Plos masculus : Calycis 

tubus subnullus ; lobi 4, triangulares, basin versus interdum denticulo aucti, vix f lin. 

longi. Corolloe tubus brevis latusque, 1^ lin. longus ; lobi 4, oblongi, 3 lin. longi, 

demum revoluti. Stamina 4, longe exserta, libera ; tilamenta gracilia, 5 lin. longa ; 

antherae pendulae, lineares, 2J lin. longye, basi bilobae, apice apiculata^. Stylus 

rudimentarius, minimus, bifidus, vix \ lin. longus. Flos femineus : Calycis tubus 

breviter ovoideus, \ lin. longus, dentibus 4, brevibus, e basi triangulari linearibus, 

aequilongis, plerumque basin versus denticulo auctis. Corolla) tubus brevis, superne 

sensim anipliatus, \\ lin. longus ; lobi ovati, aequilongi, demum recurvi. Stamina 4, 

etfeta, libera, in clusa ; filamenta \-\ lin. longa; antherai ovata), basi biloba), apice 

apiculatse. Stylus basi ad f lin. indivisus, crassiusculus, tunc bifidus; stigmata 

minute papillosa, tiliformia, flexuosa, 10 lin. longa. Bacca eUipsoidea, 2 hn. longa, 

calyce persistente coronata. 



DE. O. STAPJb^ OX THE PLORA OF MOUNT KIXABALU. 187 

Pakapaka, 10,500 feet (Maviland, 1073) ; " on the top of the mountain " (probably at 
13,000 feet) (Lota). 

Miquel deseribed a species from East Java (10,000-11,000 feet) under tlie name of 
Coprosma simdana, which may be very near to C. cm^sicauUs; however, I have not seen 
a specimen of it, and I find his description too incomplete to enable me to recoi^nize 
the species from it. Besides, he says *' folia subtus venoso-reticulata," which certainly is 
not the case in the Kinabalu plant. This belongs to a group of species which is repre- 
sented inNcAY Zealand by C.faetkUssimay Forst., and C 6v///tY//^/, Hook, lil., and in South- 
east Australia by C. nltida. Hook, fil., but without approaching to them so closely as 
the following species. 

CopnosMA HooKERi, Stapf, n. sp. Frutex humilis prostratus, ramosissimus, (jhihorrimus. 
Rami tetracjoni\ Internodia hrevia, in veterioribus circiter t-2 lin. lonira, in ramulia 
brevtssima,! lin. longa,pro maximd parte stipulis pUis minusve persistentilms ohtecta. 
Folia subsessilia vel petiolo ad 1 lin. longo tenui suffulta, ohlonga vel latweolato- 
ohJonga, 4-6 lin. longa, 1^-2 lin. lata, acuta, coriacea, nigricantia, subnitida, nervis 
veuisque inconspicuis. Stipulae lirmie, albidte, brevitcr Iriangnhiri-oraliP^ 1-H Hn. 
longa^, crasse breviterque funbriata?. Flores solifarii vel masculi bini ad apices 
ramulorum. Flos masculus : Calyx tetragono-turbinatus, breviter 4-dcntatus, 1 lin. 
longus, dentibus minute ciliatis. Corollae tubus breviter infundibuliformis, 1-1^ lin. 
longus ; lobi ovati, aequilongi, recurvi. Stamina 4, exserta, libera ; lilamentii 1\ lin. 
longa ; antherfe oblongoc, 2^-2f lin. longtp, basi biloba?, apicc npiculata*. Styli 
rudimentarii minimi 2. Flos femineus : Calycis tuhnsotoideus, 1 lin. longus ; limbus 
4-5-dentatus, dentibus triangularibus, subiequilongis. Corolla infundibuliformi- 
carapanulata, 2J lin. longa, ad medium 4-5-loba. Stamina lr-5 ; filaraenta brevia ; 
antherae eifetae, lineares, \\ lin. longse, basi bifida?, apiculata?. Styli 2, a basi liberi, 
lanceolato-acuminati, 4 lin. longi, minute papillosi. Bacca ellipsoidea, 2-2^ lin. 
longa, calyce persistente coronata. 
In crevices near the top of Mt. Kinabalu, alt. 13,000 feet [Haviland, 1072). 
Nearly allied to C. nitida, Hook, fil., from Victoria and Tasmania, and to C. ctineata, 
Hook, fil., from New Zealand, the Auckland Group, and Campbell Island, and of a very 
similar habit. But in C. nitida the branches are constantly covered with a very fine, 
almost velvety tomentum (the plant is not glabrous as stated in Benth. Fl. Austral, 
iii. 430), the internodes are longer and are soon deprived of their deciduous and very 
short stipules, and the flowers are of a somewhat difi"erent shape. C. cuueata, on the other 
hand, differs chiefly in the terete branches, very short and truncate stipules, smaller 
obovate or obversely lanceolate leaves, and very slender tube of the female corolla. 

COMPOSITE. 

Gnaphaliuji ltjteo-album, Linne, Spec. Plant. 851 ; var. multiceps. Hook, fil., Fl. 
Br. Ind. iii. 288. 
In rice-fields at Kiau, alt. 2500 feet {JBurbidge). 



188 DR. O. STAPF ON THE FLOEA 01' MOUNT KINABALU. 

Distribution (of species) : Almost cosmopolitan. 

1 am not able to distinguish GnapliaUum JungJiuhnianimi, Miq., and G. Reinwardt- 
ianum, Miq., from this form. 

PiCHEOCEPHALA LATiFOLiA, DC. in GuiU. Archiv. Dot. ii. 518. 
Bed of Dahombang River, alt. 3000 feet {Kamland^ 1274). 
Distribution : A weed, widely spread throughout the Tropics of the Old World. 



CAMPANULACE^. 

Peatia boeneensis, Hemsley in Hook. Ic. PL 1. 1532. 

, var. GEANDiELOEA, Stapf, n. var. Racemi robustiores, 2-3 poll, tantum longi, 

Calycis lobi latiores, ad 1 lin. glabri. longi. Corolla duplo major, lilacina. 

On the DahombaDg River, alt. 3000 feet (Haviland, 1309). 

Distribution (of species) : North Borneo. 

The general habit and the structure of the flower agree so perfectly with the plant on 
which the species was founded that I prefer to describe the Kinabalu specimen rather 
as a variety than as a species, though the flowers are twice as large. 

Pentapheagma aueantiaca, Stapf, n. sp. Caulis imprimis apicem versus asperulus. 
Polia petiolo 1-lf poll, longo asperulo suffulta, oblique lateque ovata, 5 poll, longa, 
3^-4 poll, lata, basi insequaliter rotundata vel subcordata, serrulata, membranacea, 
supra glaberrima, subtas asperulo-papillosa, septempli-nervia, venis venulisque 
utrinque prominulis. Cym{3e axillares, breviter pedunculatse, cernutje; bractea3 infima3 
4-6 lin. longse, late ovatse, flores arete congestos involucrantes ; interiores minores 
grosse papillostfi ; pedicelli subnulli. Calyx campamilatus, 1 poll, longus ; tubus sub- 
villosus; lobi glabrati, oblongi, obtusiusculi, 5-6 lin. longi. Corolla aurantiaca, 
^ poll, longa, e calyce breviter exserta ; segmenta late obovata, apiculata. Stamina 
circa stigma sulcatum conniventia et cohserentia 1^ lin. longa ; antherae muticae. 
At 6000 feet (Hacilcmd, 1227). 
Very well marked by the large campanulate calyx, which becomes wider after flowering 

(to \ inch in dried specimens), but not much longer, and by the orange-coloured corolla. 

There are three species described of this genus, and one or two more might be added from 

the material in the Kew Herbarium. They are distributed from the Malay Peninsula to 

New Guinea. 

Campantjmcea celebica, Blume, Bijdr. 727. 

On the side of torrents, Tampassuk Valley, alt. 2500 feet (llavlland, 1358). 

Distribution: Malayan Archipelago, eastward as far as Celebes; Indo-China to 
Chittagong and the Sikkim Himalaya. 



DK. 0. STAFF OX THE FLORX OF .MOUNT KIXABALU. 181) 

ERICACE/E. 

VACCiNirM coRiACEUM, Hook. liL, in Hook. Ic. PI. t. 892. 
At 9000 feet {Zoto ; Havihnuh 1132). 
According to Dr. Haviland, the calyx and corolla are pink. 

Vaccinium BuxiroLiUM, Hook, fil., in Hook. Ic. PL t. 891. 

Erom 7500 to 11,800 feet {Low; Haviland, 1083, 1081). 

The corolla is not always distinctly veutricose, but often tubular, as in V. coriaceum ; 
it is, however, larger than in that species (to 3 lin.). The anthers are spurred or spurless, 
and the style is slightly hairy at the base. 

Yaccinium, sp. near V. Tar'nujicefoVnim, Miq. 

A bush or small tree, at the Penokok, alt. 3000 {ao^t [Hat Hand, 1321). 

This is probably different from 1\ varingifpfolftnn, but the material is too incomplete 
for description. 

Vaccinium pachydehmum, Stapf, n. sp. Erutex erectus, inflorescentii\ ct gemmis exceptis 
glaber. Eanii crassiusculi. Eolia supernc congcsta, i)etiolo crasso 1-3 lin. longo 
suffulta, ohovata vel rotimdato-obofata, 2-4 poll, longa, 1-2 poll, lata, basi subabruptc 
vel longiuscule in petiolum attenuata, apicc late rotundata, crasse coriacca, subtus 
glandulose-punctata, margine acuto subrecurvo, nervis lateralibiis uirinque circiter 
5, angulo 30° ortis, laxe reticulatim Acuosa, nervis venisque prominulis. Raccmi 
axillares, 1^ poll, longi, cinereo-hirfelli ; bracteae oblongee vel obovato-oblongiE, 3 lin. 
longae, mox decidiiae, minute hirtella^y deinde glabrata?, ciliatce ; pedicelli curvati, 
demum 4-6 lin. longi. Calycis tubus ovoideo-hemispha^ricus, 1-1 5 lin. longus, 
hirtellus, dentibus 5 rotundato-ovatis vel ovatis^ ^-f lin. longis, ciliatis. Corolla 
(ex alabastris tantum nota) subglabra. Eilamenta pilosa ; antheras oblongte, appendi- 
cibus tubulosis, dorso minute 2-sj)inulosaB. Stylus crassiusculus, basi hirtellus. 
Bacca (submatura) globosa, 3 lin. dimetiens, vertice inter calycis dentes persistentes 
valde convexo, protruso, hirtello. 
In barren and stony open places, alt. 10,500 feet ; Maripari Spur, alt. 5500 feet, in 

flower only. [Havilandy 114.0, 1264.) 

Allied to V. arbutoides, Clarke, from the Patkoi Range in North Burma, and in a 

minor degree to V. Teysmannii, Miq., from Java, but distinct by the few nerved obovate 

leaves, and also from V. arbutoides by the short broad calyx-teeth and from V. Teysmanuii 

by the indumentum. 

Vaccinium cordifolium, Stapf, n. sp. (PI. XV. d. 7-11.) Erutex. Rami dense patuleque 
albo- vel cinereo-j)ilosi. Folia cakle approxiinata, internodia J-} poll, longa multo 
siiperantia ideoqiie laxe imbrlcata, subsessilia tel petiolo crasso 1-1^ lin. longo piloso 
suffulta, e basi auHculato-cordata omda, f-lf poll, longa, f-lj poll, lata, acuta vel 
obtusiuscula, coriacca, primu utrinque albo-pilosa, mox glabrata, supra lucida, nervis 
lateralibus 4-5, cum venulis laxe reticulantibus utrinque prominulis. Racemi ex 

SECOND series.— BOTANY, VOL. IV. 2 C 



190 -DE. 0. STAFF OX TEE FLORA OF MOUNT KINABALU. 

axillis summis, interdnm pseudo-terminales, singuli vel pauci, 2-2i poll, longi, dense 
patuleque albo- vel cinereo-hirtelli ; bract ese ovata?, 5-4 lin. longse, f oliacese ; pedicelli 
nutantes, 4-6 lin. longi. Calypis tubus albo- vel cinereo-hirtellus ; lobi late ovati, 1 J 
lin. longi, pilosi, minute ciliati. Corolla pallide rosea, glaucescens, glabra, ovato- 
tubulosa, 5 lin. longa ; lobi 5, brevissimi, late ovati, recurvi. Pilamenta e basi 
triangulari subulato-attenuata, 3 lin. longa, pilosa ; antherse oblongse, 1 lin. longse, 
appendicibus fequilongis, sparse glandulose pilosulge, postice bispinulosse. Stylus 4 
lin. longus, crassus, ad f pilosus. Ovarium 5-loculare. Bacca calyce majusculo 
coronata, globosa, 3 lin. dimetiens, alba. 
From 8800 to 9000 feet [Bcwiland, 1133, 1134, 1148). 

A very remarkable species, whicb is nearest allied to the Polynesian species Vaccinium 
cereum, Porst. The densely crowded foliage gives it quite a singular appearance, not 
unlike that of certain sj^ecimens of Gaultheria oppositifolia. Hook. fil. 

Vaccinium, sp. near V. Teysmamiii, Miq. 

At 8800 feet (Ilavilcmd, 1107). 

This is very probably a new species, but the material appears to me insufficient for 
description. The leaves are very coriaceous and not unlike those of V. Teysmannii, but 
they are almost sessile and their nervation is more prominent and of a slightly difPerent 
character ; the racemes are stiff and upright, 2-3 in. long, and the young fruits ovoid, 
crowned by short ovate and ciliate calyx-lobes, usually with a few black glands on the 
margin. I have seen but one deformed flower. 

Vaccinium micrantherum, Stapf, n. sp. Arbor humilis, inflorescentia excepta glabra. 

Kamuli graciles. Polia subsessilia, ovato-lanceolata, 2-2J poll, longa, 10 lin. lata, 

basi late cuneatim attenuata, apice longiuscule acuminata, tenuiter corlacea, nervis 

lateraiibus utrinque 2, rarius 3, fere e basi ad vel ultra medium pr or sus ductis, superne 

circiter 3 utrinque additis, angulo 30*^ ortis, reticulato-venosa, nervis venisque leviter 

prominulis. Eacemi ex axillis superioribus orti, 2-pollicares, graciles ; bractege cito 

decidual, e cicatricibus probabiliter minutae ; pedicelli 3 lin. longi, recurAd vel patuli, 

plus minusve hirtelli. Calycis tubus breviter cupularis, vix J lin. loDgus, cinereo- 

hirtellns ; lobi late triangulares, ciliatuli. Corolla ovato-tubulosa, 3 lin. longa, ore 

parvo, intus inferne sericeo-pilosula ; lobi minute erecti, minute ciliolati. Filamenta 

\ lin. longa, pilosa ; antherse breviter lineari-oblongse, J lin. lowjcc, appendicibus tubu- 

iosis dimidio brevioribus, postice minute bispinulosae. Stylus crassiusculus, ad 

medium hirtellns, 2J lin. longus. Ovarium vertice hirtello. 

At 3500 feet (Haviland, 1296). 

Allied to V. Sasseltli, Miq., but differing considerably in the small anthers and the 
short filaments, and also in the hairy calyx and the much thinner leaves. 

Gaultheria borneensis, Stapf, n. sp. (Pi. XV. c 4-6.) Pruticulus, 2 pedes altus, 
giaberrimus. Ramuli graciles. Polia subsessilia, obovato-oblonga vel breviter 
obianceolata, 4-5 lin. longa, 1^-2 lin. lata, basi et apice acuta, margine utrinque sub 
recurvo, minute glanduloso-denticulata, denticulis 4-6, coriacea. Racemi ex axillis 




DE. O. STAFF ON THE PLOEA OF MOLM Kl.XAEALU. 191 

superioribus vel summis orti, snepe pseudo-termiaales, 1-lh poll, loiif/l; bracteu^ late 
ovatae, vix 1 lin. lougcc, diu persistentes ; pedicelli ceruui, ad medium vol su[)i-a bibrac- 
teolatl.Z im.lon^i. Calyx fere ad basiu o-partitus; segmonta ovata, | lin. longa. 
Corolla alba, ovato-tubulosa, 3 lin. longa, dcntibus porrectis minutis. Filameiita 
llin. longa, linearia; antherae ovatiB, 1 lin. longae, asperulae ; loculi poro apicali de- 
hiscentes, uterque appendioibus 2 setaceis, reflexis, patulis oriiatus. Ovarium globo- 
sum, 5-lobatimi, ut stylus giabrum. Ca[)su]a baccata, all)a. 

At 12,000 feet (Havllcuid, 1085). 

Very similar in habit and foliage to Gaitltheria antlpoda, Forst., but distinct by the 
comparatively long racemes, the bibracteolate pedicels and larger flowers. 

DiPLYCosiA CHUYSOTHiiix, Stapf. (PL XIV. E. 10-11.) Prutex epipliytieus. Rami 
graciles, dense pllis aureo-rufis demum cinerascentibus aeloso-hu'snll. Folia pctiolo 
\-^ lin. longo suffulta, ovata, 8-10 lin. longa, 5-6 //;/. lata, basi rotundata, apice 
acuta, coriacea, subtus plerumque leviter coneava, supra glabcrrima, lucida, subtus 
et imprimis in margine setosa, nervis utrinque 1-2, supra impressis, siepius indistinctis. 
Flores solitarii vel pauciiii folioruni axillis; pedicelli densissimesctoso-hirsuii, 2-3 lin. 
longi, apice bibracteolati ; bracteobe parva), cupulatim connatie. Calyx fere ad basin 
5-partitus ; segmenta ovata, 3 lin. longa, acuta, exlns densisaime sericeo-hirsuta^ intus 
minute adpresseque pilosula. Corolla rosea, late ciqndari-camj^amUata, 5 lin. longa, 
glabra; lobi late ovati, 1 lin. longi. Filamenta glabra, 11 lin. longa; anthene 
ovatae, minute asperulae, appendicibus rectis, paulo brevioribus, oblique apertis. 
Ovarium disco angusto cinctum, globosum, 5-lobatum, minute pilosulum ; stylus 
4 J lin. longus. Placentae prope loculi basin ortae, multi-ovulatae. 
At 6600 feet {Haviland, 1182). 

DiPLYCosiA RUFA, Stapf, n. sp. (PL XIV. d. 8-9.) Frutex epiphyticus. Rami dense 
patnleqiie ritfo-setosi, setis demum cinerascentibus. Folia brevissime petiolata, ovata, 
1^2| poll, longa, 8-11 lin. lata, basi rotundata, acuminata vel acutissima, coriacea, 
supra glaberrima, nitida, infra et imprimis in margine rufo-setosa, nervis lateralibus 
utrinque paucis, basalibus plerumque supra distinctis. Flores axillares, solitarii 
vel pauci, fasciculati, e gemmis perulatis orti ; pedicelli denmssime setosi, 2-3 lin. 
longi, apice bibracteolati ; bracteolae cupulatim connatae, extus dense setosae. Calyx 
fere ad imam basin partitus ; segmenta lanceolata, 3-3| lin. longa, ex^tiis denslssime 
pilis aiireo-rufis sericeo-hirsuta, intus glabra. Corolla angulato-campamdata, 3|-4 
lin. longa, ad f 5-lobata ; tubus inferne et in commissuris glaber; lobi ovati, acumi- 
nati, extus rufo-hirsuti. Stamina ssepe post coroUam delapsam persistentia ; filamenta 
IJ lin. longa, glabra ; antherae ovatae, 1 lin. longae, minutissime asperulae, appendici- 
bus rectis aequilongis, apice poro oblique apertis. Ovarium disco lobulato cinctum, 
globosum, 5-lobatum, liirtellum ; stylus 2 lin. longus. 

From 7500 to 8000 feet (Haviland, 1137, 1181). 

This and the preceding species are allied to D. pilosa, Blume, hut very distinct in the 
dense and long shining indumentum of the bracts, pedicels, and calyces, and the coarser 

2c2 



192 I>E. O. STAPi^ ON THE FLOEA OF MOUNT KINABALU. 

hairs of the branches and leaves. Both species certainly are closely allied, but sufficiently 
distinct by the size and shape of the leaves and the shape and hairiness or glabrousness of 
the corolla. In both cases the flowering and the vegetative branches originate from 
perulate buds. The bud-scales are much smaller in Biplycosia chrysothrix, more 
hirsute, and almost hidden by the bristles of the stem. 

DiPLYCOSiA ciLiOLATA, Hook. fil. in Hook, Ic. PI. t. 894. 
Erom 6000 to 8000 feet {Loio ; Hcwiland, 1135). 

DiPLTCOSiA MTETiLLTis, Stapf, u. sp. Prutcx. Ramuli graciles, novelli parce setulosi, 
mox glahri. Polia petiolo 1-1^^ lin. Ion go suflfulta, elliptico-oblonga, 5-7 Un. longa,S 
lin. lata, utrinque acuta vel apice niucronulata, in margine incUstincte crenulata vel 
Integra et parce setulosa, setulis mox deciduis, coriacea, supra lucida, subtus glandu- 
loso-punctata, cost^ supra impressa, subevenia. Plores solitarii, rarius bini, axillares; 
pedicelli hirtelli, 2 lin. longi, basi minute bracteati, apice bibracteolati ; bracteolse 
ovatse, basi connatce, ^-f lin. longae, ciliolata^. Calyx ad medium 5-lobus, 1^ lin. 
longus ; lobi ovati, acuti, ciliolati. Corolla campanulata, 3 lin. longa, fere ad medium 
5-loba ; lobi apice recurvi. Pilamenta e basi ovata abrupte linearia, glabra, I5 lin. 
longa ; antherse ovato-lineares, J lin. longse, asperulse, in appendices dimidio breviores, 
apice poro minuto apertos, abeuntes. Ovarium disco cinctum, globosum, apice 
hirtellum, cseterum ut stylus 3 lin. longus glabrum. 
At 8700 feet {Haviland, 1136). 

DiPLYCOSiA PiNiFOLiA, Stapf, n. sp. (PI. XIV. A. 1-3.) Erutex glaberrimus. Kami 

nigro-fusci, graciles, dense foliati. Polia juvenilia fere filiformia, exisccando 

nigricantia, adulta anguste linearia, 2-2J poll, longa, f-1 lin. lata, basi in petiolum 

brevissimum attenuata, obtusiuscula, integerrima, coriacea, enervia. Plores axillares, 

solitarii vel geminati ; pedicelli tenues, 2-3 lin. longi, apice bibracteolati ; bracteolse 

rotundatse, ciliolatoe. Calyx 1 lin. longus; segmenta 5, ovata, acuta vel subacumi- 

nata, albo-ciliolata, apice recurva. Corolla (juvenilis) calyce vix longior, glabra. 

Disci glanduli 5, rotundi, ovarium cingentes. Pilamenta glabra ; antherse ovatse, in 

appendices tubulosos rectos abeuntes. Ovarium ovoideum, sparse pilis longis 

vestitum ; stylus glaber ; stigma minute bilobum. 

At the Penokok, alt. 3500 feet {mwiland, 1297). 

This is a species of very peculiar habit, if compared with the remainder of the genus, 
whilst it approaches nearest to D. ciliolata. Hook, fil., in its flowers. 

DIPLYCOSIA HETEROPHYLLA, Blume, Bijdr. 853 (?) ; (forma parvifolia, an species nova ?). 
At 7600 feet (Haviland, 1146). 

I have seen only fruiting branches. Some of the specimens of D. heterophylla (for 
instance Beccari, PI. Sum. n. 243, and in a less degree a specimen collected by Lobb on 
Mt. Salak in Java), have, at least partly, leaves which approach those of Haviland's plant 
very closely, except that the latter are still more acuminate. Tlie nervation, which is hardly 
visible in the Kinabalu specimens, seems to be of the same type as in D. heterophylla, 




BE. 0. STAFF ON THE FLORA OF .AIOUNT XIXABALU. ] 1);J 

whilst the fruits agree perfectly. For these reasons I am ratlun- inclined to regnrd the 
plant enumerated as a small leafy form of the variahle Dipft/nos/'a hrferophylla ; but I 
must add that it differs from all specimens of D. heteroplnjUa whicli I haA'c seen in the very 
slender branches, the long acuminate leaves, and the solitary or sometimes paired flowers. 

PiPLYCOSiA PUNCTTTLATA, Stapf, n. sp. Frutcx, inflorescentia exceplj\ glaber. Ramuli 
crassi, pallide badii, minute nigro-glauduloso-punctati. Folia petiolo crasso t-0 I'ui. 
longo suffulta, clliptica, 2J-3 poll, longa, li-1 J poll, lata, basi breviter attenuata, 
apice acuta vel subacuminata, margine acuto subrecurvo, vakle coriacea, ntnuqne 
glanduUs nigris punctata, nervis plerumque ntrinqne angulo 70° oriis, infra mar- 
ginem arcuatim collectis, supra impressis, infra prominulis, venis paucis. Flores 
4-6, in fasciculis axillaribus et extra-axillaribus ; pedicelli basi brncteA. parva sulTulti, 
hirtelli, ad 2 lin. longi, apice 1)ibracteolati ; bracteolye hirtelhe, late ovatte. Calyx 
2 lin. longus, glaber, nigricans, ad medium vel ultra 5-lobatus ; k)bi ovati, minute 
glanduloso-ciliolati. Corolla in alabastro angulato-ovata. Filamenta glabra; 
antherae lineari-oblongse, minute asperulae, appendicibus multo brevioribus longitu- 
dinaliter fere ad basin apertis. Ovarium disco cinctum, globosum, 5-lobatum, ut 
stylus glabrvim. 

At 7000 feet (Ilavilaml 1202). 

Also allied to I), heterophylla, Blume, and J), scahrida, Becc, but characterized by the 
longer petioles, the more numerous lateral nerves, and the copious glands, particularly of 
the young branches. 

DiPLYCOSiA PENDTJLirLORA, Stapf, n. sp. (PI. XIV. c. 7.) Frutex glaberrimus. Rami 
graciles, pallidi. Folia petiolo crasso, supra sulcato, 8 lin. longo suffulta, elUptica 
vel ovato-elliptica, 6-7 ^jo^/. longa, 2|-4 ^jo//, lata, basi rotundata vel subcordata, 
abrupte acuminata, coriacea, iiifra nigrc-punctata, nervis lateralibus utrinque circa 
4-5 supra impressis, infra prominulis, in nervum a raargine remotum collectis, venis 
paucis teuuibus. Flores solitarii vel gemini, e ramnlis orti, pcndnJ'i ; pedicelli bract eis 
minutis suffulti, 1 poll, longi, fiViformes^ apice bibracteolati ; bracteolae ovatae, basi 
connatse, J-J lin. longge. Calycis fructus semimaturi tubus piriformis, carnosulus, 
2 J lin. longus ; limbus 5-partitus ; lobi ovati, minute glanduloso-ciliati. Capsula 
vertice glabra ; stylus IJ lin. longus. 
On Kinitaki River, alt. 5000 feet {HavRand, 1269). 
Though the material is incomplete, I do not hesitate to describe this very remarkable 

species, the affinity of which lies evidently with B. heterophylla, Bl. It is strongly 

marked by the large leaves and the slender and long pedicels. 

DiPLYCOSiA KiNABALUENSis, Stapf, n. sp. (PL XIV. B. 4-6.) Frutex prostmtus, pattcos 
pollices alius. Rami crassiusculi, glabri, teretes. Folia petiolo crasso, 2-3 lin. longo, 
suffulta, elliptica vel obovata vel oblonga, 2^-2f poll, longa, IJ-li poll, lata, basi acuta, 
apice rotundato-obtusa, margine arguto subrecurvo, setoso-ciliata vel demum gla- 
brata, coriacea, opaca, 5- vel rarius 7-plinervia, nervis infimis 2 fere ad apicem ductis, 



194 JDE. O. ST^PF OiS TEE FLOEA OF MOUiS'T KINABALU. 

2 sequentibus sub ipso apice cum costa conjunctis, nervis superioribus venisque 
utrinque impressis, grosse reticulantibus. Flores extra-axillares, iu ramulis annotinis 
solitarii vel pauci, fasciculati e gemmis perulatis; pedicelli 3-4 lin. longi, glabri, 
basi bracteati, apice bibracteolati ; bracteolae rotundato-ovatse, 1 lin. longae, ciliatse. 
Calyx late cupularis, fere 2 lin. longus, fere ad basin 5-partitus ; segmenta late 
ovata, glabra, ciliata. Corolla campanulata, virescens, 4J lin. longa, ad f 5-lobata ; 
lobi ovati, brevissime obtuseque acumiuati, porrecti. Filarnenta 2-2| liu. longa, in 
marginibus asperula ; antherse lineari-oblongae, IJ lin. longae, antice leviter incurvai, 
asperulae, in appendices tubulosos, apice oblique apertos, J lin, longos attenuatae. 
Ovarium disco cinctum, globosum, glabrum ; stylus glaber, superne incrassatus, 2J 
lin. longus. 

From 10,500 to 12,000 feet (mwiland, 1086). 

According to Dr. Haviland the flowers are close to the ground. 

DiPLYcosiA mcEOLATA, Stajif, n. sj). Frutex, inflorescentiis exceptis^ glaberrimus. Folia 
petiolo crasso 3-6 lin. longo suiFulta, ohovata vel ohovato-ohlorga, 2-3 poll, longa, 
1-1^ poll- lata, hasi interdum Jonge in petiolum cdtenuata^ apice rotundata, margine 
acuto subrecurvo, coriacea, supra lucida, infra crebre nigro-punctulata, 5-plinervia, 
nervis infiniis fere marginantibus versus apicem evanidis, sequentibus 2 sub apice 
ipso cum costa connexis, supra impressis, infra prominulis, caeterum subevenia. 
Flores axillares et extra-axillares, in ramis annotinis 3-6 fasciculati ; pedicelli hirtelli 
vel glabrati, 2-3 lin. longi, basi minute bracteati, apice bibracteolati ; bracteolse 
rotundato-ovata^, ciliolat^, J lin. longae. Calyx sub anthesi late cupularis, deinde 
paulo accrescens, turbinatus, ad J 5-lobatus, IJ-lf lin. longus; lobi late ovati, 
ciliolati. Corolla urceolata^ vix 2 lin. longa, 5-dentata, dentibus late ovatis, 
marginibus revolutis. Filamenta glabra, |- lin. longa; antheras ovato-oblongae, 
filamentis eequilongae, asperulaj, in appendices brevissimos late apertos abeuntes. 
Ovarium globosum, ut stylus, glabrum. Capsula semimatura carnosula, turbinata, 
fere 3 lin. longa. 
At 8000 feet (Haviland, 1150). 

DiPLYCOSiA MEMECYLoiDES, Stapf, n. sp. Frutex, inflorescentiis exceptis, glaber. Rami 
teretes, subgraciles. Folia petiolo 2-3 lin. longo suffulta, late ovato-elliptlca, 
2 poll, longa, 1-lJ poll, lata, basi rotundata vel hrevissime attenuata, abrupte 
acuteqiie acuminata, margine subplano, coriacea, exsiccata nigro-fusca, utrinque 
lucidula, epunctata, 5-7-plinervia, nervis 2 infimis usque ad acumen, sequentibus 

2 fere ad apicem productis, tenuibus, prominulis, casteris tenuissimis, rectis, utrinque 
4-7. Flores axillares et extra-axillares, solitarii vel 2-3; pedicelli 2-3 lin. longi, 
liirtelli, demum glabrescentes, basi minute bracteati, apice bibracteolati ; bracteolae 
ovata), ciliolatse, basi connatse, f lin, longae. Calyx sub anthesi late cupularis, 
\h lin. longus, ad \~% 5-lobatus ; lobi ovati, ciliolati. Corolla tidmloso-campanulata, 

3 lin. longa, ad ^ 5-lobata ; lobi late ovati. Filamenta glabra, I5-I J lin. longa ; 
antberse ovato-oblongae, minute asperulae, | lin. longae, in appendices dimidio 



DE. 0. STAPF OX THE FLORA OF .AIOUXT KIXABALU. 195 

breviores, apice oblique apertos, attenuata}. Ovarium disco anp^usto cinctum, 
globosum, apice pilosum ; stylus glaber, superne incrassatus, vix 2 lin. longus. 
At 6000 feet {Havllaud, 1208). 

DiPLYCOSiA CINNAMOMIFOLIA, Stapf, n. sp. Frutcx ^lab(n'. Hamuli exsiccjiti, sub- 
angulati. Folia j)etiolo 3-4 lin. lungo crasso sufFulta, ovata vel obloni^o-ovata, 
2-^-4-2- poll, longa, 1^-2 poll, lata, basi rotundata vel leviter cordata, snishn et 
S(Spius longiuscitle acuminata^ coriacea, supra sublucida, subtus nigro-punctata, 
5-pliiiervia, nervis lateralibus 2 infimis ad f, sequentibus 2 fere ad apicem 
ipsum productis, csetei am fere evenia. Flores e ramulis annotinis extra-axillares, 
3-6 fasciculati; pedicelli basi minute bracteati, birtelli, 2-2i lin. longi, apice 
bibracteolati ; bracteola? late ovatjB, basi connatoe, ciliolatae. Calyx turbinatO' 
campanulatus, glaber, vix 1^ lin. longus, ad \ vel \ 5-lobatus; lobi rotundato-ovati, 
ciliolati. Corolla rosea, ureeolaris, 2 lin. longa^ sub ore constricta, dentibus ovatis 
brevibus. Filamenta f lin. longa, glabra ; antherie sequilonga), ovata?, minute 
asperula?, appendicibus brevissimis late apertis. Ovarium disco cinctum, globosum, 
5-lobatum, ut stylus glabrum. 
From 8000 to 9000 feet {Ilanhaid, 1138). 

Rhododendron Lowii, Hook. fil. in Ilook. Ic. PI. t. 883. 

From 8000 to 10,500 feet {Zow, Burbldge\ Haviland, 1081). 

Distribution : Owen Stanley Hange, New Guinea, according to Sir F. v. Mueller. 

The capsules are 1-1^ in. long and up to \ in. broad. The valves separate down to the 
base and come off at last. They are comparatively thin, but hard and woody, with a 
broad, pellucid, and wavy margin. Mr. Burbidge says in a note that the flower- 
clusters measure 12-14 in. in diameter, and that the shrub attains a height of from 
10 to 30 feet. 

Ehododexdron CRASSiFOLirM, Stapf, n. sp. Frutex epiphyticus, glaber. Rami 
crassiusculi, cortice cinerascente. Folia probabiliter dimorpha in innovatione 
quaque, inferiora alternantia delapsa cicatrices parvas relinquentia, summa 2 
subopposita vel 3 inflorescentias subtendentia et innovationes novellas in axillis 
foventia, petiolo crassissimo 4 lin. Ion go suffulta, oblonga, 5-7 poll, longa, 2-J-3 
poll, lata, basi auriculato-cordata, apice obtusiuscula, valde coriacea, glabra, cost^ 
valid^, supra prominente, subtus canaliculata. Rami floriferi brevissimi, crassi, 
dense cicatricibus pcrularum ol)siti, pauciflori. Pedicelli If poll, longi. Calyx 
obsoletus, diseiformis. Corolla cdba riibro-suffusa, superne campanulata, |-1 poll, 
longa, ad vel paulo ultra medium 5-lobata ; tubus brevis, cylindricus, intus, basi 
viUosula excepta, glaber ; lobi obovato-oblongi vel oblongi. Stamina 10 ; filamenta, 
e basi incmssata villosula, filiformia, 5 lin. longa; antherse exsertai, breviter 
obovatJT, 1 lin. longie. Ovarium glabrum, breviter cylindricum, 5-loculare, in 
stylum fere | breviorem attenuatum ; stigma dilatatum. 
On Kinitaki River, alt. 5000 feet [Low). 



196 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU 

Allied to Bhododendron Lowii, but the flowers reddish white and much smaller. 
There are 4 flowers with the specimen, 3 of them detached. It appears to me, from the 
scars on the axis of the inflorescence, that there were not more on it. 

Ehopodendhon Bkookeanum, Low, ex Lindl. in Journ. Hort. Soc. iii. (1848) p. 42. 

*' A lanky shrub, growing on a tree-fern, but taller than it," at 7800 feet (Haviland, 
1149). 

Distribution : Xorth Borneo. 

Rhododekdrox stenophyllttm, Hook. fil. ex Beccari, ' Malesia,' i. 203. Burbidge, 
' The Gardens of the Sun,' 275, with fig. on p. 274. Prutex glaber, 2-4 ped. 
altus. Ptami gracilcs. Polia dimorpha, persistentia, intervallis 1-2 poll, longis 
congesta, numerosa (ad 15), partim subverticillata, angustlssime lUieari-lan- 
ceolata vel linearia^ li-2f poll, louga, 1-2 lin. lata, basi in petiolum brevis- 
simum attenuata^, acuta vel interdum fere setaceo-acutata, coriacea, sparsissime 
lepidota, uninervia ; decidua pauca ad basin interstitiorum, pauca alia sequaliter 
inter persistentia distributa, submembranacea, scariosa, vel superiora in folia 
persistentia abeuntia, lanceolata, acuta vel subulato-acuminata. Inflorescentia 
terminalis e gemmis perulatis aj)hyllis, joauciflora, ut videtur plerumque 2-flora. 
Ferulae circiter 6-8, subsequilongae, ovatai, 6-8 lin. longse, glabrse, tenues. 
Pedicelli crassiusculi, 3 lin. longi, glabri. Calyx obsoletus, disciformis. Corolla 
coccinea, superne campanulata, 1 poll, loriga, ad fere medium 5-loba ; tubus brevis, 
latmsculus ; lobi obovati. Stamina 10 ; filamenta 5-6 lin. longa, basi pilosula ; 
antherse oblongse, 2 lin. longte. Ovarium cylindricum, 2J lin. longum, dense 
albo-tomentellum, 5-loculare ; stylus 3 lin. longus ; stigma dilatatum. 
In very shady, mossy places, from 5000 to 8000 feet {Low, Burbidge; Haviland, 1104). 
Closely allied to M. salicifolium, Becc, but diifering in the very narrow leaves and 
smaller corollas with a wdder tube. The specimens collected by the three gentlemen 
mentioned are very uniform. The sequence of the leaves is very remarkable, though 
typically the same as in the other species of § Eu-Uhododendy^on. Sir Hugh Low's 
specimens as well as Haviland's were collected in the spring. At this time all branches 
terminate with a cluster of persistent leaves which correspond to the " Sommerblatter " 
of B. ponticum (Maximowicz), and surround a very young bud. To judge from the 
dried material, these buds grow out during the following season into young branches of 
1-2 m. in length, shifting the inner bud-scales along with them, till at a certain time the 
normal leaves appear in great numbers, terminating tlie growth of that season. What 
the length of that period is I do not know, but it seems very probable that more 
than one shoot is formed every year. The branches may grow for several consecutive 
periods without lateral ramifications, but when these are formed they always spring from 
the axils of '• summer leaves," and in a very limited number (2-1). The membranaceous 
or scarious leaves pass into the summer leaves sometimes through one or two inter- 
mediate leaves, but more often they follow each other abruptly. Different as R. sleuo- 
phyllimi may seem if compared with H. Javanicum, tirny yet belong to the same group. 



mi. O. STAPF 0.\ THE FLOEA OF .MOUM KINABALU. 197 

Rhododendrox eugosum, Low, ex Hook. fil. in Hook. Ic. PI. 1. 885. 

From 7000 to 10,000 feet (Zoic, Jiurhidgc ; BavilancU 1139). 

The capsules are cylindric, about 6 lin. long-, and densely pubescent. The valves 
separate to the base, but they do not twist or recurve. 

Khododendron acuminatum, Hook. fil. in Hook. Ic. PI. t. 886. 
Erom 6000 to 8000 feet (Low). 

Ehododendron dueionifolium, Beccari, ' Malesia,' i. 202. 

Terrestrial or epiphytic, from 4000 to 7000 feet (Low, Burbidye ; Haviland, 11G3, 1208). 

Distribution : jSTorth-west and North Borneo. 

A specimen collected by Low, and named B. polj/anthinn in MS. bv Sir Joseph 
Hooker, has some leaves much larger than most of the other specimens have, the largest 
measuring 7 in. by 2J, but ther(» is no other difference. One of Dr. Haviland's 
branches has leaves of an intermediate size. 

Rhododexdeon lacteum, Stapf, n. sp. (§ Eu-Rhododcndron), (PI. XV. a. 1-2.) Frutex 
epiphyticus. llami dense cinereo-iomenlello-hirtelli. Polia alternantia vel verti- 
cillata, sessilia, linear i-ohlonga, l-lf poll, longa, J-J poll, lata, basi rotujidata, 
apice obtusa, margine recurvo, coriacea, supra minutissime A/He//«, sublucida, subtus 
in costd longius, cceterum brevissime sjiarseqiie hirtellay subuninervia. Inflores- 
centia terminalis, e gemma perulata aphylla, multitlora. Perulai minute sericeo- 
tomentellai. Pedicelli 3-4 lin. longi, sericeo-hirtelli. Calyx obsoletus. Corolla 
alba, hypocrateriformis ; tubus subrectus vel rectus, lJ-2 poll, longus, viUosulns ; 
limbus 5-partitus ; lobi oblongi, 8 lin. longi, 3-3^ lin. lati, glabri. Stamina 10 ; 
filamenta tenuissima, ad 3 lin. exserta; antherse ovat9B,jf lin. longa?. Ovarium 
cylindricum, dense sericeo-pilosum, 3 lin. longum, 5-loculare ; stylus sericeo-pilosulus, 
sub apice glaber, 2 poll, longus ; stigma dilatatum ; lobulatum. 
Penokok, alt. 3500 feet (Haviland, 1295). 
Allied to H. jasminiflonmi, Hook, fil., but very distinct by the narrow leaves and the 

indumentum of the branches, leaves, and flowers. 

E/HODODEXDEON VERTiciLLATUM, Low, cx Lindlcy, Joum. Hort. So3. iii. (1S4S), 86; 
Hook. fil. in Hook. Ic. PI. t. 884. 

Epiphytic or terrestrial on rocky banks of torrents, from 8000 to 11,500 feet (Xt^/^?; 
Savilarid, 1147). 

Distribution : North-west Borneo (Mt. Penrisen) and North Borneo. 

Ehododexdeon buxifolium, Low, ex Hook. fil. in Hook. Ic PI. t. 890. 

Pound only at 12,000 feet, evidently very rare (Loic ; Baciland, 1080). 

The capsules are oblong, ^ in. long, and sparingly covered with smaU scales. Both 
collectors state that they found it on one place only, and both give the elevation at 
12,000 feet. 

SECOXD SEEIES. — BOTAXY, VOL. IV. 2d 



198 DB. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

Rhododendeon cuNEiFOLirM, Stapf, 11. sp. (§ Fragrantes). (PI. XY. b. 3.) Arbor 
humilis. Eamuli graciles, novelli badii, demum albo-cinerascentes. Polia petiolo 
1 lin. longo suffulta, ohovato-ciineata, 6-8 lin. longa^ siqjerne 2 lin. lata, basi longe 
cuneatim in petiolum attenuata, apice rotuiidata, emarginata, margine leviter 
reciirvo, impresso-punctata, coriacea, utrinque sparsim lepidota, supra demum 
glabrata, lucida, uninervia, costa supra impressa. Flores terminales soJitarii vel 
bini e gemma perulata, ovoidea, acuta, 2 lin. longa. Ferulae paucse, ovatse, 
exteriores breviores, sericeo-ciliatulae. Pedicelli filiformes, 2 lin. longi, dense 
lepidotL Calyx breviter 5-lobatus, lepidotus. Corolla rubra, tubulosa, superne 
sensim leviterqii^e dilatata, 5 Im. longa, extus lepidota ; lobi 1-1^ lin. longi, 
porrecti (an semper ?), ovati, margine suberosuli. Stamina 10 ; filamenta filiformia, 
glabra, breviter exserta ; antberae ovato-globosae, \ lin. longse. Ovarium dense 
lepidotum ; stylus tenuis, glaber, 3 lin. longus. 
At 7000 feet {Raviland, 1180). 

Allied to M. Vidalii, Kolfe, R. apoanum, Stein, It. rosmarimfolium, Vid., and H. 
quadrasianuni, Vid., but distinct by tbe very narrow cuneate leaves and the rather tubulose 
corolla. Of these I know R. rosmarinifolium only from Yidal's description ; I have 
seen the other three. According to Vidal, R. rosmarinifolium differs by robust 
branches, oblong linear leaves, smaller bell-shaped flowers, and hairy pedicels. R. 
apoanum is a native of Miudanao, R. quadrasicmum inhabits South and Central Luzon, 
whilst R. Vidalii {=R. verticlllafiim, Vidal, non Low) and R. rosmarimfolium were 
found in North Luzon. They all come very closely together. 

Ehododendhon ericoides, Low, ex Hook. fil. in Hook. Ic. PI. t. 887. 

Prom 8000 feet almost to the top, very common at 10,500 feet {Low ; Jlavilana, 
1066) : in sandstone grit and peat {Biirhidge). 

Prom 6 to 10 feet high at the lower limit, dwarfed (6 in.) at the higher elevations. 

Clethua canescens, Reinwardt, in Blume, Bijdr. 863. 

In young jungle, Ulu Tawaran, alt. 2000 feet (Havilcmd, 1367). 

Distribution : Java, Lombok, Celebes, North Borneo (Kulawat, alt. 2000 feet, Lohb 1). 

The species is represented by very closely allied forms in South China (6*. Faberi, 
Hance) and in Luzon iC. lancifolia, Turcz.). 

EPACBIDE^. 

Leucopogon malayanus. Jack in Mai. Misc. I. v. 20. 

Maripari Spur, alt. 5500 feet (Haviland, 1 263). 

Distribution: Malaya, on the mainland as far as Malacca, in the island of Phukok 
in South Cambodja, and from Sumatra to Borneo. 

LErcopoGON stjaveolens, Hook. fil. in Hook. Ic. PL t. 898. 

Prom 7500 to 10,000 feet (Low) ; from 9000 to 13,000 feet (Ilaviland, 1062). 
Distribution : North Borneo, Mindanao. 



DE. 0. STAFF ON THE TLOIM 01- MOUNT KIXABALU. 19i) 

Tlie leaves vary from linear oblong- (3-1.^ by 1^ liii.) to linear (5 by 1 lia.) ; but the 
venation is always the same. 

M^SA INDICA, Wallich, in Uoxb. El. Ind. ed. Carey Sc Wall. ii. 230 in nota. 
Trom 4000 to 5000 feet (Haviland, 1191, 1192, 1301). 
Distribution : Tropical Asia. 

Myrsine capitellata, Wall, in lloxb. El. Ind. ed. Carey & Wall. ii. 295. 
, var. AVENis, Clarke in Hook. lil. El. Br. Ind. iii. 512. 

Erom 8000 to 11,500 feet (Mcwilaml, lODO, 1151). 

Distribution (of speeies) : Himalaya, from Nepal to Bliotan, Khasia Hills, Silhet, 
Assam, Malay Peninsula and Archipelago, Hainan, Ceylon, and Western Peninsula to 
the xAnamaily Hills. 

The Himalayan form, on which the species was founded, is very uniform in every 
respect, and some of the Khasia specimens approach it very closely, whilst others pass 
more or less into the common Malayan form, which was named M. avenls by De Candolle. 
Exactly the same veinless form occurs in Ceylon and in tiie Western Peninsula, whcn-e the 
more luxuriant specimens or low-region forms, however, are not inclined to develop leaves 
of quite the same shape as they do in the Khasia Hills, but a form which is more 
oblanceolate or obovate and which occurs in stunted specimens of the higher regions 
of the Khasia Hills. Quite similar forms occur in Borneo and the Philippines and 
in Hainan, but wherever the leaves are enlarged they maintain their very coriaceous 
consistency, the nerves never become so prominent as in the Himalayan or the 
common Khasia form, and the leaves are always widest in the upper third or fourth. The 
shape of the calyx-lobes and the length of the pedicels vary too much to afford a 
tixed character lor specific distinction. The same seems to be the case with regard to 
the tetramery or pentamery of the flowers, as in the specimens from Kinabalu the 
flow^ers are 4- or 5-merous. Therefore, I think, M. affinis, DC, 31. Hasseltil, Blume, and 
M. bonieensls, Scheif., must be merged into M. capitellata v. avenis, and I doubt very much 
whether M. densiflora, Scheff., of which I have seen only a barren branch, can be kept up. 

The Kinabalu specimens have exceptionally large fruits (3 by 2^ lin.) and very 
obtuse calyx-lobes. The flowers are always pedicelled and less numerous than they 
usually are in the variety avenis. 

Myesine dasyphylla, Stapf, n. sp. Erutex glaberrimus. Rami crassiusculi, cicatri- 
cibus foliorum et ramulorum floriferorura residuis obtecti, cortice nigricante, plus 
minusve late sulcati, superne, ut ramuli, denskslme foUatl. Eolia subsessllla, 
spathulato-cuneata, f-f poll, longa, 4-5 lin. lata, obtusa, hand emarginata, crasse 
coriacea, supra glaiico-vlrklia, lucida, subtus opaca, glandulose punctata, uninervia, 
aveuia. Elores pauci vel solitarii, sessiles, in ramulis axillaribus brevissimis, gemmis 
perulatis. Perulee bracteaeque minute glandulose ciliata?. Calyx f lin. longus, 
ad f 4-5-lobatus ; lobi obtusi, late ovati, minute papilloso-ciliati vel subglabri. 

2d2 



200 DE. O. STAPr ON THE ELORA Or MOUNT KINABALU. 

Corolla 1 J lin. longa ; tubus brevissimus ; segmenta oblonga obtusa, demum reflexa, 
in niargine intusque joapillosa. Antlierae sessiles, basi coroUae segmentis adnatse, 
^ lin. longse. Ovarium ovoideum, apice annulatuni ; stylus ^ lin. longus, irregu- 
lariter lobulatus. Fructus globosus, apiculatiis^ pallidus, 2 lin. dimetiens ; stylus 
diu persistans. 
At 11,000 feet (Hcwilcmd, 1065). 

Nearly allied to M. capitellata v. cwenis, Wall., and to M. simiatrana, Miq., but 
distinct from the former by the exceedingly dense foliage, the glaucous, sessile or 
subsessile leaves, and the sessile flowers ; from the latter by the dense foliage, the leaA^es, 
which are not emarginate and not pellucid-dotted, the sessile flowers ; and from both by 
the more apiculate fruits. The dense foliage is a very striking character, and has no 
parallel even in the most stunted specimens of M. capitellata I have seen, nor in 
those of M. sumatrana. 

Myesine, sp. alfinis M. achradifolice, F. Muell. 

Maripari Spur, alt. 5500 feet {Haviland, 1258). 

There is but one branch with fruits in the collection. The stem and the leaves are 
exceedingly like those of M. achradifolia from Queensland, the only difference being in 
the petioles, which are 4 lin. long, in the Kinabalu plant, and 1 inch in M. acJiradifolia. 
The flowers are pedicelled as they are in this species, and the calyx also is very similar, 
but, perhaj)s, a little smaller. The fruits are comparatively small, almost depressed, 
globose, about IJ-lf lin. in diameter. As I do not know the flower of Dr. Haviland's 
plant, and as the fruit of M. acJiradifolia is unknown, I abstain from describing the 
Kinabalu plant as a new species. 

Labisia pothoixa, Lindley, in Bot. Eeg. 31, t. 48. 
From 3000 to 1000 feet (Burbidge). 
Distribution : Malay Peninsula, from Singapore to Penang ; from Sumatra to Borneo. 

E31BELIA TOETIOSA, Stapf, n. sp. Frutex glaberrimus. Bami divaricati, exsiccati plus 
minusve irregulariter sulcati, cortice nigricante. Folia petiolo crasso 1-2 lin. longo 
suffulta, oblonga vel obovato-oblonga, l|-lf ^joZZ. longa, ^jwll. lata, basi rotundata, 
apice obtusa, rarissime subacuta, crasse coriacea, exsiccata supra fusco-olivacea, 
subtus ferruginea, nervis lateralilms temiissimis numerosis, scepe indtstinctis, angulo 
75°-90^ ortis. Paniculge terminales in ra?imlisfoliatis 3-4 poll, longa?, 2|-3 poll, latce, 
rohnstiores, strictse, subebracteatae. Pedicelli 1-2 lin. longi. Calyx ad vel ultra 
medium 5-lobus, 1\ lin. dimetiens ; lobi ovati, obtusi vel acutiusculi, minute 
ciliolulati. Fructus globosus, vix 2 lin. dimetiens, styli basi persistente, ajjiculatus. 

At 8800 feet (Haviland, 1113). 

A very well-marked species, allied to U. Javanica, A. DC. 

Embelia coriacea, a. DC. in Trans. Linn. Soc. xvii. 135. 
From 7600 to 8800 feet [Hamland, 1152). 



DE. O. STAFF OX THE FLOUA OF MOUNT KINABALU. 201 

Distribution : Malay Peninsula as far north as Ponang; Arcliipelago from Sumatra to 
Borneo. 

Beccari's specimens, nos. 767, 878, 1189, 3922, all from Sarawak, belong also to 
this species. 

Embelta PH.EADENIA, Stapf, n. sp. Erutex. Pami ramulique gracilcs, novelli dense 
ferrugineo-hirtelli, deminn 2>h(s winusve (jJahmti, cortice pallide badio. Folia 
petiole 2 lin. longo suflfnlta, ovata vel ovato-oblonga, I -\\ poll. Jonga, interdum 
multo minora, basi rotunda vel breviter lateque acutata, apice obtuse acuminata, 
rarius obtusa, sub apice pauci-crenata vel suhintegra, firme membranacea, pallide 
viridia, glanduloso-punctata, glandulis luce permissd nihiginosia, nervis lateral ibus 
primariis utrinque 5-8, secondariis intorjectis cum venulis in reficulationiun 
gracillimam, densissimam, utrinque prominulam intertextis. Racomuli axillares, 
paueiflori, saepe 1-2-flori, 2 lin. longi ; bractejc minuue, ferrugineo-furlunicoa'. 
Pedicelli gracillimi, 1\ lin. longi. Calyx fere ad basin 4-partitus, J lin. dimetiens ; 
segmenta ovata subacuta, glandulosa et furfuracea vel glabraf a. Friictus ellipsoidec- 
globosus, 1| lin. longus, i)allidc badius, sub vertice rubiginoso-glandulosus ; stylus 
filiformis, 4 lin. longus, diu persistens. 

At 0000 feet {Ilavilaud, 1203). 

A very fine and distinct species, allied to E. Gardneriana, Wight, and E. Myriilliis, 
Kurz. The glands are most conspicuous in transparent light. 

Embelia minutifolia, Stapf, n. sp. Frutex. Pami ramulique dense ferruginei. 
Folia petiolo gracili 1 lin. longo vel breviore suffulta, obovata, 2 lin. lonya, \\ lin. 
lata, basi breviter cuneata, apice rotundata et minute cremilata, coriacea, glabra, 
minute punctata, subenervia. Pacemuli axillares, valde redacti, saipe uniflori ; 
rliacMs ferruginea, brevissima, bracteata ; bractese congestse, minutse. Pedicelli 
graeiies, 1 lin. longi. Calyx fere ad basin 4-partitus ; segmenta ovata, sub apice 
glanduloso-punctata, praetereaque minute gland uloso-ciliolata, 3 lin. longa. Corolla 
ignota. Ovarium ovoideum, sparse glanduloso-pilosulum, in stylum Ij-plo longi- 
orem abiens. Bacca globosa, j)aulo ultra 1 lin. dimetiens. 
At 8500 feet {Haviland, 1117). 
Closely allied to E. Myrtillus, Kurz, but differing in the smallness of all parts and 

the shape of the leaves, probably also in the habit, for it seems to be of a stiff(M' and 

spreading growth. 

Embelia sptr^oides, Stapf, n. sp. Frutex glaherrimus. Pami graeiies, novelli badii, 
demum cinerascentes, lenticellis albidis conspersi. Folia petiolo 3 lin. longo 
suffulta, elliptica, vel ovato- vel obomto-elUptica, f-lj poU. longa, 6-8 lin. lata, 
basi acutata, apice obtusa vel rotundata, rarius subacuta, a medio ad apicem crenato- 
dentata, coriacea, supra olivacea, subtus pallidiora, subglauca, et punctulis minutis 
pra?tereaque glandulis majoribus luce transmissa vix rubiginosis notata, nervis latera- 
libus utrinque 7-8, cum venis venulisque in reticulationem densam, gracillimam, 



202 DK. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

utrinque leviter prominentem intertextis. Racemi axillares o-12'flori, rhachi 2-4 
lin. louga ; bractese minutce. Pedicelli graciles, 2 lin. longi. Calyx fere ad basin 
4-partitus ; segmenta ovata, ^ lin. longa, glanduloso-punctata, hand ciliolata. 
Corolla fere ad basin 4-partita ; segmenta obovato-oblonga, 1 lin. longa, glanduloso- 
punctata, miuntissime ciliolulata. Stamina corollam suba^quantia. Ovarium sparse 
glanduloso-papillosum, ovoideum, in stylum paulo longiorem abiens. Bacca 
globosa, apiculata, 2^ lin. dimetiens, rubra, crebre glanduloso-punctata. 
At 11,500 feet (ITaviland, 1089). 

This species belongs evidently to the same group as the two preceding species. It is 
very distinct by its glabrousness. The prefloration varies. The more frequent case 
is, one segment innermost, the other three overlapping to the left ; then I found one 
segment outside, the opposite odc inside, the lateral ones overlapped by the former and 
overlapping the latter. But in one case I observed a fifth smaller segment, and in this 
case there were two segments outside, two inside, and the fifth inside, overlapped 
perfectly by an outer and an inner segment. Upon these facts the prefloration hardly 
afi'ords a good character for subdividing the genus. 

Akdisia oocarpa, Stapf, n. sp. Arbor humilis, glaberrima. Bamuli superne ancipites. 
Folia petiolo crasso, alato, 3-1 lin. longo suffulta, lanceolata, 8-10 poll, longa, 2-2 J 
poll, lata, basi in petiolurn decurrentia, apice subacuminata, tenuiter coriacea, supra 
nigro-fuscescentia, nitida, subtus rubro-brunnea opaca, elevatim punctata, nervis 
lateralibus utrinque 25-30, angulo 50°-60° ortis, supra leviter iaipressis, subtus 
prominulis, indistincte venulosa. Panicula terminalis, laxa, interdum foliata, 5-7 
poll, longa; rami inferiores racemose ramificati, superiores vel fere omnes simplices, 
umbeiligeri, sursum curvati. Pedicelli demum semipollicares. Calyx fere ad 
basin 5-partitus; segmenta late ovata, acutiuscula, 1^ lin. longa. Corollas segmenta 
ovata, acuta, 2 lin. longa. Pilamenta brevissima ; antherse 1| lin. longa?, acuta?. 
Stylus 2 lin. longus. Bacca ovoidea vel ovoideo-globosa, apiculata, 3J-4 lin. longa, 
longitudinaliter striata, basi calyce arete adpresso circumdata. 

Penokok Biver, alt. 3000 feet (Havilcmd, 1317). 

Very similar in habit to A. paniculata, Boxb., and A. colorata, Boxb., but very distinct 
by the large ovoid berries. 

Akdisia brachythyksa, Stapf, n. sp. Prutex glaberrimus. Rami robusti, novelli 
exsiccando nigrescentes, demum cinerascentes. Eolia subsessilia vel jietiolo brevls- 
simo crasso suffulta, oblonga vel elliptica, 3-5 poll, longa, 1^-2^ poll, lata, basi 
\)reviter attenuata vel subrotundata, brevissime acuteque acuminata vel acuta, 
margine subrecurvo, valde coriacea, supra lucidula, subtus minute nigro-punctata, 
nervis lateralibus angulo 70°-80° ortis, primariis utrinque 14-20, secundariis inter- 
jectis, arcuatim sub margine coliectis, nervo marginali alio extero irregulariter 
undulato vel arcuatim fracto addito, grosse laxeque reticulata, nervis venisque 
utrinque prominulis. Panicula terminalis brevis, 2>ollicaris, cermia, ramis umbelli- 
feris vel corymbuliferis, 3-5 lin. longis. Pedicelli circa 3 lin. longi, recti vel s^epius 



DR. O. STAVF 02i THE PLOEA OF MOUAT KINABALU. 1 203 

recurvi. Calyx ad vel ultra medium 5-lobus, 1^ liu. dimetiens; lob i triangulares, 
mimite ciliolati. Corollae segmenta orata, aristato-acurainata, 2 lin. longa ; alabas- 
trum acutum. Eilamenta brevissima ; authernp subsagittatfr, apiculatto, 1 lin. longae. 
Stylus 1^ lin. longus. Bacca subdepresso-globosa, 2^ liu. dimetiens. 
At 3000 feet (?) (Eatlland, 1386). 

I have not been able to identify this plant with any species in our Herbarium ; but as 
we have no complete set of the Malayan species, as enumerated in Scheffer's * Commen- 
tatio de Myrsinaceis/ I at first hesitated to describe it as a new species. However, the 
short, nodding, and truly terminal inflorescence, and the shortly petioled or almost 
sessile leaves, seem to distinguish it from any species which otherwise might come into 
consideration. Amongst these are Ardisia Icevigata, Blume, A. dingiensis, O. Kuntze, 
and particularly A.javanica^A. DC. 

AnDisiA JAVANiCA, A. DC, Prodr. vii. 130. 
At 8800 feet {Haviland, 1144). 
Distribution : Malaya, from Sumatra {Beccarl, 400) to Borneo. 

Ardisia amabilis, Stapf, n. sp. Prutex glaberrimus. Hamuli graciles, subanfracti, 
lineis 2 elevatis notati. Eolia pctlolo 1-1^- Un. longo suff'ulta, ovato-lauccolafd, 2J-3i 
poll, longa, 13-15 lin. lata, basi subrotundata vel breviter acutata, apice acuminata, 
subcoriacea, supra nigro-fuscescentia, opaca, elevatim punctata, subtus ferrugineo- 
fusca, nigro-punctulata, costa utrinque prominente, nervis lateralibus 10-12 tenuissi' 
mis, angulo 60° ortis, evenia. Panicula terminalis, gracilis, nutans, 3 poll, longa ; 
rami racemose ramificati ; ramuli infimi 1-2-flori ; superiores umbelligeri, corymbose 
dispositi. Pedicelli 2-3 lin. longi, graciles. Calyx fere ad basin 5-partitus 1 lin. 
dimetiens ; segmenta ovata, acuta, minutissime ciliolata. CoroUse segmenta ovata, 
caudato-acuminata, IJ lin. longa, deflexa, rubro-glandulose-punctata ; alabastrum 
acutum. Stamina erecta ; filamenta brevia ; antherse sagittatse, apieulatse, vix 1 
lin. longse. Stylus antheras ^ superans. 
At 6000 feet {Kamland, 1211). 
Allied to A. Gardneri, C. B. Clarke, and to A. andamannica, Kurz, but distinct by the 

very short petioles, the shape of the leaves, the very indistinct nervation, and the small 

calyx. 

AuDisiA viRENS, Kurz, Por. PI. Burma, ii. 575. 

Shady jungle at Kiau {Burhidge) ; Penokok Eiver, alt. 3500 feet {Haviland, 1315). 

Distribution : Khasia Hills and Assam to South-west China ; Borneo. 

The berries are scarlet, and measure 4 lin. in diameter. The pedicels are slender, 9-14 
lin. long, whilst the peduncles measure 4-15 lin. in Dr. Haviland's, and 6-12 lin. in 
Burbidge's specimens. The inflorescence is truly terminal. The leaves are crenulate, 

with "-lands between the crense. This character, as well as the truly terminal inflores- 

f? 

cence, disagrees with the description of A. vireus, Kurz, as given in Hooker's ' Plora of Brit. 
India,' but the very specimens quoted by Mr. C. B. Clarke also possess crente and truly 



201 DE. 0. STAPP ON THE PLOEA 0¥ MOUNT KIXABALU. 

terminal inflorescences, though sometimes on short lateral and leafy branches. The 
fruits of the Bornean specimens are larger than any of ArcUsia vireus I have seen, but the 
difference is evidently within the range of individual or racial variation. A. undulata, 
Clarke, is partly at least identical with A. mrens. Partly it seems to merge into A. 
crenulata, Roxb., whilst on the other hand A. macrocarpa, Wall., approaches also very 
much to A. virens. Yet, as A. crenata is very uniform over a wide area, including 
Malaya, Tropical China, and Japan, it seems to me better to retain it as a separate species, 
though closely connected wiWx A. virens, which may or may not be merged into A. macro- 
carpa, and which is found from the Kliasia Hills over Manipur and Assam to South- 
west China, and in a hardly distinguishable form in Borneo. 

Akdisia villosa, Roxb., M. Ind. ed. Carey & Wall. ii. 274, forma glabrata. 
On the Kadamaian, alt. 4500 feet {HavilancU 1242). 
Distribution : Malaya, on the mainland as far as Mergui. 

Ardisia oxyphylla, Wallich, Cat. n. 2291 ; A. DC. Prodr. viii. 130. 

At 5000 feet {Raciland, 1212). 

Distribution : Malaya, on the mainland as far as Tenasserim, Java, Borneo (Ceram '?). 

I doubt very much whether this species can be separated from A. polycephala. Wall., 
which seems to be only a form with a stouter and richer inflorescence. In that case, the 
area would include also Burma up to Bamu and the Shan States. 

EBENACEiE. 

DiosPYROS, sp. afF. D. aurecB^ Hiern, Trans. Cambr. Philos. Soc. xii. pt. i. 20G. 6 . 

At 3000 feet (Haviland, 1322). 

This is very probably a new species, differing from i>. aurea by the longer flowers, 
which are caudate in the bud on account of the very tightly twisted lobes. The tube is 
\ in. long, and but little longer than the lobes, which are reflected in the open flower. I 
found in the single good flower which I examined 14 glabrous stamens, arranged in two 
bundles of 3 and four bundles of 2 stamens each. 

DiOSPYROS, sp. 

Small tree, on the Dahombang, alt. 3000 feet [Haviland, 1314). 

This species is in habit very like Diospjyros pergamena, Hiern. The flowers spring, 
according to Dr. Haviland, from near the base of the stem. But, as they are all deformed, 
I must refrain from describing or naming the plant. 

STYEACE^. 

Symplocos adenophylla, Wallich, Cat. n. 4427 ; C. B. Clarke, in Hook. fil. Plor. Br. 
Ind. iii. 575. 
Maripari, 5500 feet to 7000 feet {Ilavlland, 1187-1256). 
Distribution : Malaya, on the mainland to Penang, Java, Banca, Borneo. 
S. iteophylla, Miq., and S, hancaua, Miq., are synonyms of S. adenophylla, Wall. The 




DE. O. STAPF ON THE TLORA OF MOUNT KINABALU. 205 

species is represented in the Kew Herbarium from Borneo by the following specimens:— 
Sarawak, Lohh ; Beccari, 2115 ; and Lobon^ Peak, Lobb. 

Symplocos l^tevieidis, Stapf, n. sp. Trutex glaberrimiis. Ilaiiii -raciles, corticc 
badio. Eolia petiolo vix 1 lln. longo subtus canaliculato suffulta, oblonga, 2-3 poll, 
longa, 1-1^ poll, lata, basi plus minusve rotundata, longiusculc acumiuata, basi 
excepta argute serrata, subpergamacea, bete viridia, nervis lateralibus utriiupie 
circa 8, sub margine arcuatim collectis, uervo marginali alio arcuatini i'racto 
addito, nervis venisque utrinque leviter prominulis. Elores subsessiU^s, in racemis 
axillaribus, gracilibus, paucifloris, j-1 poll, longis. Fructus ellipsoideus, purpureas, 
4 lin. longus, calycls lobis ovatls obtusis, submembranaceis, 1 lin. longis, coronatus. 
Embryo subrectus. 
At the Penokok, alt. 4000 feet {UacUand, 1320). 

Yery similar to S. ramoslsslma, Wall., from the Himalaya, but the leaves are very 
shortly petioled and more serrate, and the fruits are cro\\ned by rather large calyx- 
teeth. 

Symplocos deflexa, Stapf, n. sp. Frutex. Rami ramulique dense* fulvo-hirsuti vel 
hirtelli, demum glabrati, nigricantes. Folia petiolo 1 lin. longo, dense hirtello- 
tomeutello suffulta, ovata vel ovato-oblonga, li-2 poll, longa, f-1 poll, lata, basi 
rotundata, acute acuminata, margine revoluto serrulate, subcoriacea, supra, costa 
pilosula excepta, glaberrima, exsiccata sa^pe olivaeeo-fusca, subtus in nervis et in 
margine necnon in parenchymate fulvo-hirsuta, nervis lateralibus utrinque 5-7, supra 
tenuibus, subtus prominulis. Eacemi paucitlori, in ramulis lateralibus terminales, 
folio summo breviores ; prseterea hinc inde flores solitarii ex axillis foliorum 
horumce ramulorum additi ; rhachis cum pedicellis brevissimis hirtella. Fructus 
ovoideo-oblongus, 4 lin. longus, basi hirsutulus, ca^terum glabratus, calycis lobis 
adpresse hirtellis vel glabratis, 1 lin. longis, coronatus. Embryo rectus. 
At 8800 feet {Hai'ilancI, 1105). 

Symplocos zizyphoides, Stapf, n. sp. Frutex. Eami ramulique dii-aricati, leviter 
anfracti, primo adpressissime griseo-flavido-hirsuti, demum giabrescentes, brunnei. 
Folia petiolo J-1 lin. longo, plerumque curvato, suffulta, ovato-oblonga, rarius ovato- 
lanceolata, 1-lJ poU. longa, J poll, lata, basi rotundata, acuminaic, margine revo- 
luto, serrulata, subcoriacea, supra exsiccando oHvacea, glaberrima, subtus in nervis 
et in margine, nee non in parenchymate, adpresse fiavido-griseo-hirsuta, nervis late- 
ralibus utrinque 5, subtus tantum distinctis, tenuibus, prominulis. Flores solitarii, 
axillares, in ramulis lateralibus. Pedicelli adpresse hirtelli, 4-5 lin. longi, sub flore 
2-3-bracteolati ; bracteola? minuta?. Calyx 1-1^ lin. longus, adpresse hirtellus, ad 
medium lobatus; lobi rotundato-ovati, ciliati. Corolla alba, 5 lin. dimetiens. 
Fructus anguste ovoideo-oblongus, 5 lin. longus, calycis lobis exceptis glabratus. 

At 10,500 feet {Saviland, 1088). 

The inflorescence is, to speak strictly, a one-flowered raceme, but the racemose cha- 
racter is only slightly indicated by the minute bracteoles. The divaricate branches, which 

second series. — botany, vol. IV. 2e ■ 



206 DR. O. STAFF ON THE FLORA OF MOUNT KTNABALTJ. 

are spread out in one plane and flexuose in zigzag-, remind one of the habit of Zizyphus 
Spina Christi, Z. Lotus, &c. 

Symplocos Johniaxa, Stapf, n. sp. Prntex. E-ami ramulique densissime subpatuleque 
ferrugineo-hirsuti. Eolia petiolo 1 lin. Ion go, hirsuto suffulta, ovata, 1-2 poll, 
longa, 7-12 lin. lata, basi rotundata vel subcordata, /^r^ c<2^«c?«^o-«<?^<m^^«f cf, margine 
piano vel basin versus recurvo, serrato-dentata, subcoriacea, flavido-viridia, supra, 
costa hirsntnla excepta, glaberrima, lucidula, subtus in nervis necnon in margine 
(infoliis noYcUis densissime) hirsuta, nervis lateralibus utrinque 5-7, supra impressis, 
subtus prominentibus, 1-1 J lin. a margine eximie arcuatim collectis, supra evenia. 
Flores solitarii, axillares^ cermd; pedicelli brevissimi, hirsuti, bibraeteolati. Calyx 

1 lin. longus, hirtellus ; lobi rotundati. Corolla alba, 4 lin. dimetiens. Eructus 
purpureus, ovoideo-oblongus, 4 lin. longus, calycis lobis exceptis plus minusve 
glabratus. 

At 6000 feet {Savilcmd, 1161). 

These three latter species are nearly allied to each other, and at the same time also to 
S. minor, Clarke, from Ceylon, which they represent in Borneo. They all differ from 
S. minor by the acuminate leaves, and S. Johniana also by the almost sessile solitary 
flowers. S. dcjlexa approaches S. minor indeed very closely, and there is hardly any 
difference apart from the shape of the apex of the leaves. S. zizyp)hoides deviates more 
iji the habit as Avell as in the complete reduction of the racemes to solitary flowers. 

Symplocos buxieolia, Stapf, n. sp. Prutex, 8 ped. altus, glaberrimus. Hami ramu- 
lique robust!, densissime foUati, nigricantes. Folia petiolo 1-1 J lin. longo suffulta, 
obovato-elliptica, 6-7 lin. longa, 4-5 lin. lata, basi breviter attenuata, apice obtusa, 
ssepe rotundata, a medio denticulata, margine prsesertim basin versus recurvo, 
coriacea, supra lucidula, olivaceo-fusca, costa impressa, subtus flavido-viridia, nervis 
lateralibus utrinque circiter 3 tenuibus prominulis. Elores solitarii in axillis 
foliorum summorum. Pedicelli (racemi rhachis reducti) ad J poll, longi, superne 
pauci- rel plurihracteati; bractea^ subfoliacese, obovatae, vel summae 3 florem sub- 
tendentes, plerumque lineari-oblongse, 2-1^ lin. longse, subpersistentes. Calyx fere 

2 lin. longus ; lobi rotundato-ovati minutissime ciliolati. Corolla 6 lin. dimetiens. 
Frucius lineari-oblongus, 6 lin. longus, calycis lobis coronatus. 

Highest parts of Kinabalu, alt. 13,000 feet (Low). 

A very remarkable species, allied to S. hractealis, Thw., from Ceylon, from which it 
differs by the smaller, very crowded leaves, and the reduced inflorescence. 

APOCYNACE^. 

Leuconotis a:nceps (?), Jack, in Trans. Linn. Soc. xiv. 121, t. t. tig. 2 (an ejus 
varietas ?), 
In young jungle on the Ula Tawaran, alt. 2000 feet (Hai'iland, 1365). 
Distribution : Sumatra, Borneo. 
In the Sumatran specimens of L. anceps the calyx-lobes are a little lono-er, the bracts 



DE. O. STAFF OX THE FLOEA OF MOOT KIXABALU. 207 

are also longer and more acute, and the inflorescence is more puberulous and richer. 
But these differences do not appear to me sufficient to justify the specific separation of 
Dr. Haviland's plant, of which I saw only one branch. 

WiLLrGHBEiA TIRMA, Blume, Mus. Bot. i. loJ^.— Syn. IT. Burhldgei, Dyer, in Kew 
Eeport, 1880, pp. 44-46. 

Near Koung- and Kiau {Burhldge). 

Distribution : Malaya, from Borneo to Sumatra, and in the Peninsula as far as Perak. 

Burbidge's specimens in the Herbarium are marked " Lawas lliver," but there is a 
sketch and a note in his journal which make it ])retty certain that the same species is 
also found at the base of Kinabalu. He states that it is a large climber, with 
golden yellow fruits and rather common, and that the Gutta hunters seem to obtain 
plenty of produce, as in most of the houses he was shown bundles of balls, some old and 
dry, having been kept two to three years, which the natives know very well improves them. 

Raitwolfia serpentina, Benth. in Benth. et Hook. fil. Gen. PI. ii. GJJ7. 

, var. GKACiLis, Stapf, n. var. Cymae 4-5-floriE, graciliores, laxiores. Calycis lobi 

anguste lineares, elongati, Ih lin. longi. 

At the Penokok, alt. 3000 feet {Ilamland, 131G). 

Distribution : Tropical Himalaya to Tenasserim ; Ceylon to Travancorc ; Java. 

Exactly of the habit of Jl. ser])entina ; bul the inflorescence and the shape of the calyx- 
lobes, which I find very uniform in the Indian and Javanese specimens, differ slightly. 

AiiTXiA LrciDA, Wall, in Boxb. PI. Ind. ed. Wall. & Carey, ii. 510. 

, var. MEIANTHA, Stapf, n. var. Corollae tubus 4i lin. longus, i. e. fere duplo 

longior quam in Wallichii exemplis. 

At 7500 feet {Sariland, 1185). 

Distribution (of species) : Malaya, on the mainland to Perak, Sumatra, Java, Borneo. 

A. qwmcUa, Miq., from Sumatra, is identical with A. lucida, Wall. 

Alyxia, sp. 

At 5000 feet (Haviland, 1184). 

The material is too imperfect for determination. Dr. Haviland says it is a straggling 
shrub with white flowers. 

Paeamekia glandtjlifeea, Benth. in Benth. et Hook. fil. Gen. Plant, ii. 715. 

, var. philippinensis, Stapf.— (Syn. P. iMlipinnensis, Radlk. in Sitz. Ber. Bayr. 

Ak. Wiss. xiv. 518.) 
At Kiau, alt. 2500 feet {HavUcmd, 1359). 

Distribution (of species) : Malaya, on the mainland to Martaban and South Cambodja, 
Andamans to the Philip23ines. 

The differential characters as indicated by Eadlkofer do not hold good when tried on 
a larger material. Hairy pits occur also on the leaves of the type specimen, the inflores- 

2e2 



208 DE. O. STAFF ON THE FLOE A OF MOUNT KINABALU. 

cence is occasionally lax in the Indian plant, and calyx and corolla vary witliin the 
extremes which are said to he characteristic of the two species. But there is a differ- 
entiation in Parameria glcmdulifera (sensu ampliore) : the Philippine and the Kinahalu 
plants are very minutely hairy in the young branches and inflorescences, just as in Radl- 
kofer's description, and they have smaller calyces than Wallich's specimens. Yet a 
Bornean specimen, collected by Motley near Banjermassin, shows the same tomentum, 
though in a less degree, w^hilst it is otherwise indistinguishable from the typical form ; 
on the other hand, a Javan specimen with glabrous panicles and branches has 
calyces hardly larger than those of Badlkofer's P. philijjpmensis. Thus I think it more 
expedient to enumerate the eastern form as a geographical variety of P. gla?icluUfera. 

ASCLEPIADACEiE. 

HoTA COKONAEIA, Blume, Bijdr. 1063. 
Near Koung and Kiau {Biirbidge). 
Distribution: Malay Peninsula ; Sumatra; Borneo, Banjermassin (Jib^/(?y, 871). 

LOGANACE^. 

Btjddleia asiatica, Lour., El. Cochinch. p. 72. 

Kadamaian Biver, alt. 2500 feet {Saviland, 1352). 

Distribution : Throughout India, Indo-China and Malaya, as far as Timor and the 
Philippines. 

GENTIANACE^. 

Gentiana lycopodioides, Stapf, n. sp. (§ Chondrophyllum.) Perennis (?). Caules 

graciles, abbreviati et parce ramificati vel elongati et subsimplices, e basi prostrata 

ascendentes, nigricantes, ad 4 poll, longi. Eolia internodiis duplo et ultra longiora 

ideoque densa et interdum densissime imbricata vel internodia sequantia et magis 

dissita, sessilia, ovata, acutissima, ssepe aristulata, crassiuscula, squarrosa, suhplicata, 

apicfi Tccurza, marginibus anguste chartaceis, dorso plus minusve carinata, \-\\ lin. 

longa. Elores solitarii, in ramorum apicibus. Calyx ccmipamdatus, 2-2J lin. longus, 

lobis tubo sequilongis, e basi ovata triangulari-lanceolatis acutissimis. Corolla cam- 

pamtlata, ad 4 lin. longa, albida, extus vittis latis olivaceis ornata, lobis brevissimis 

apiculatis, plicis apice latis, rotundatis, minute crenulatis. Eilamenta infra medium 

dilatata. Ovarium late oblongum, l|-lf lin. longum, bremter lateque stipitcitum, 

angustissime marginatum, stigmatibus subsessilibus. Capsula (submatura) ellip- 

soidea, inchisa, 2i lin. longa. 

In boggy, sheltered, but sunny places, at 11,000 feet (IlamUuid, 1060). 

G. lycopodioidcs belongs, like G. borrieensis, Hook, lil., to the multifarious group of 

G. quadrifaria, BL, which is still very little understood. It is chiefly cliaracterized 

by the comparatively large and bell-shaped flowers and the peculiar habit. The leaves 

— even those at the base of the stems — are very small and rigid. The stalk of the capsule 



DK. O. STAFF ON THE FLORA OF MOUNT KINABALU. 209 

is very short. It apj^roaches most to Gentlana Ettingshausenii, F. v. Muell., from the 
Owen Stanley Range in Xew Guinea. 

Gentiana boeneensis, Hook. fil. in Hook. Ic. PI. t. 1110 B. 

Temhurnngo, alt. 7500 feet (Havilmid, 1160 ; Loic). 

From the more complete material now in my hands, this species is doubtless 
specifically distinct from G. quadrifaria, Blume, thou^^h allied to it. 

SOLAN ACE ^. 
SoLAXUM BiFLORUM, Lour., Fl. Cochinch. 129, not of Don. 

On Kinitaki Eiver, alt. 4000 feet {Low) ; alt. 5000 feet (JIaviland, 1234). 

Distribution : Throughout Tropical Asia. 

It seems to me very probable that .S'. subtriincatum. Wall., S. bigeminatum, Nees, 
S. l(Bi'e, Dunal, S. deuticulatum, Blume, S. macrodoii. Wall., ^S*. crassiprtahnu. Wall., and 
probably also S. himense, Miq., and 8. nematoscpuJnm, Miq., are only forms of one verv 
variable species which was first described by Loureiro, under the name of S. hiflormn. 
The supposed differentiul characters which are derived from the shape o!" tlic leaves, the 
indumentum, the presence or absence and the length of the calyx-teeth arc in reality verv 
inconstant, and they hardly allow the distinction of geographical subspecies or varieties. 
The Kinabalu ^ilant is very similar to the form described as *S'. la'}:c by Dunal, ])ut the 
linear calyx-teeth rise from below the calyx-limb. They agree also in all characters with 
*S'. Blumei, Nees, but for the presence of calyx-teeth. Froiri typical S. denticidatmn they 
differ chiefly in the glabrous leaves, and from 8. hiflormn. Lour., as it is usually represented 
by the Chinese specimens, in the glabrous or almost glabrous calyx and shorter, less acute 
teeth. Some of the Khasia specimens, named S. crassipetalum, Wall., and S. macrodon, 
Wall., agree almost exactly, or they differ solely in the length of the teeth or the presence 
of a more copious tomentum. 

BORAGINACE^. 

Havilaxdia, Stapf, n. gen. 

Calyx profunde o-fidus, segmentis oblongo-linearibus, post anthesin accrescentibus, 
tenuiter membranaceis, venulosis. Corolla hypocrateriforinis, tubo brevi, cylindraceo, 
recto, fauce gibbis 5 papillosis instructa; lobi 5, imhricati, obtusi, patentes. Stamina 5, 
tubo afhxa, inclusa, filamentis brevissimis ; anthenie ovato-oblongae, obtusae. Ovarii 
lobi 4, distincti, gynobasi planse impositi ; stylus filiformis, brevis, stigmate depresso- 
capifato. Nucules 4 vel abortu pauciores, erectae, oblique ovoideas, suhtnriuetra:, 
ventro acute carinatw, angnlis lateralihm obtiisis, basi smpe brevlter stipitatce, la^vis- 
simfe, atrse, pericarpio tenuiter crustaceo, cum areola basilari rainuta gynobasi affixse. 
Semina oblique ovoidea ; cotyledones planae. Herba multicaulis prostrata vel repens, 
strigillosa. Folia alterna. Flores albi, ad axillas solitarii. 

Havilandia bokxeensis, Stapf, n. sp. (PI. XVI. a. 1-9.) Caules longe prostrati 
vel repentes, interdum radicantes, parte inferiore ob folia emarcida longiuscule 



210 DE. O. STAPF OX THE FLOEA OF MOUNT KINABALU. 

deniidati, superne parce divaricatim ramosi, disticlie foliati, adpresse cinereo-strigil- 
losi. Polia sessilia, semiamplexicaulia, ob basin cauli arete adpressam breviter 
semiyagmantia, spathulato-cuneata, leyiter emarginata, 7-10 lin. longa, 2-3 lin. lata, 
in margine superne adpresse, basin versus patule ciliata, supra sub apice adpresse 
strigillosa, subtus, costa excepta, glabra, uninervia. Elores pedicello 1-2 lin. longo, 
gracili, adpresse strigilloso suffulti. Calycis tubus brevis, demum 1 lin. longus ; 
lobi demum 2 lin. longi, setuloso-ciliati, circa nuculas matnras plus minusve patentes. 
CoroUse tubus -5 lin. longus; lobi tubo sequilongi, rotundati, basi constricti. 
Antherfje ^ lin. longce. Stylus J lin. longus. Xaculae vix 1 lin. longce. 
At 11,500 feet {Haviland, 1059) ; at 11,000 feet, in a small marsh [Low). 
Savilandia comes nearest to Myosotis, particularly to those extreme forms which were 
described as M. antarctica by Sir Joseph Hooker, and M. spathulata, Eorst. M. spathu- 
lata is found in jS^ew Zealand, M. antarctica in jS^ew Zealand, Campbell Island, and in 
Magellan Strait. In both species the flowers are solitary in the axils of normal leaves, 
but sometimes the upper leaves are more or less reduced, thus forming an approach to 
the leafless inlloresceDce of most species of Myosotis. In Uavilandia the upper leaves do 
not, as a rule, bear flowers ; but they do so occasionally, and, so far, there would hardly be 
su£&cient reason to separate the Bornean plant generically from Myosotis. The two 
New-Zealand species above mentioned, however, have exactly the same nutlets as the 
typical Myosotis of the Northern Hemisphere, This is not so in Uavilandia. Here 
they are triquetrous, with an acutely carinate ventral edge, and very blunt lateral edges. 
This is the main character on which the new genus is founded. The prefloration of the 
corolla of Myosotis is stated to be always contorted. If this w^ere really the case, another 
differential character might be found in the fact that it is distinctly imbricate in Savi- 
landia. But I would not lay too much stress upon that. First, the way in which the lobes 
overlap each other is not constant in Uavilandia, and then M. spatJmlata and M. antarctica 
at least are exceptions from the type of aestivation in Myosotis. The usual form in Uavi- 
landia seems to be exactly the same wdiicli is figured by Eichler in the diagram for 
Echiiim vulgare (' Blnthcndiagramme,' 197). But I find also that sometimes the lobes 
which correspond to the lateral lobes of the lower lip of Ecliium are overlapped on both 
sides, in consequence of which the lobe between the calyx-segments 2 and 4 is free 
on both sides. Tiie flowers of Echimn are slightly zygoniorph. Avery slight irregularity 
may be found sometimes also in Uavilandia, but normally the corolla is regular. In 
M. spathulata I found in the few buds I examined the same aestivation as in J^chium, 
whilst in M. antarctica the other form, described above for Uavilandia, seems to be more 
common. 

The new genus approaches, on the other hand, Trigonotis, which also has imbricate 
prefloration, but very characteristic nutlets of a different shape. 

SCBOPHULABINE^. 

Euphrasia bohneensis, Stapf, n. sp. fPl. XYI. b. IO-ICj.) Herba perennis, a basi 
interdum lignescente multiramosa, 3-10 poll, alta, glabra, vel superne minutissime 



DR. O. STAPF ON THE FLORA OF MOUNT KIXABALU. 211 

piiberula, exsiccanclo nigricans. Eolia obovato- vel oblongo- vel ovato-cuneata, circa 2 
lin. longa, in petioliim brevissimum attenuata, subobtusa, utriiique obtuse 2-i-crenata, 
margine revoluto. Flores soUtar/i hi axilUsfoUonim seciiniltmi totam fere loiKjitu- 
dineni ramulorum hornotinorum ; pedicelli graciles, vix 2 lin. longi. Calyx 2-2] lin. 
longus, ad medium 4-fidus ; lobi lincares, obtusissiuii, in marginibus et prceci2)H(? 
apice incrassati. Corolla alba, flavo-maculata ; tubus vix exsertus ; lal)iuni inferuui 
2^-3^ lin. longum, laciniis cuneatis truncatis retusis, lateralibus intermedio paulo 
brevioribus; labium superum subgaleatum, subcEquilongum, ad ^ bilol)ULn, lol)is 
obtusissimis rul)ro-striatis. Antlieme colia3rentes, subglabra?, basi bilobue, lobis mar- 
gine pilosulis, mucronatis. Stylus pilosulus. Capsula caJfjce paulo hrevlor, obocoidea, 
obtusissima, apice pilosula vel glabra. 
From 8000 to 12,000 feet {Maviland, 1061). 
Nearly allied to JSuj^/irasla collina^ II. Br., from Australia, but distinct chiefly by the 

solitary axillary flowers and the short obovato capsule. Tlie New-Zealand species are 

also allied. 

BiiooKEA ALBICANS, Stapf, n. sp. Rami novelli dense brevisslmeque lanafo-tomentosi, 
demum glabrati, teretes. Polia petiolo 1-2 poll, longo suffulta, ovata, 3-8 poll. 
longa, lJ-3 J poll, lata, basi acuminata, in petiolum decurrentia, acuminata, serrulata, 
supra primo farinosa, mox glabrata, nigricantia, subtus Incano-tomentosa, nervis 
lateralibus utrinque 10-13, cum venis laxe reticulantibus supra imprcssis, infra 
prominentibus. Racemi axillares et terminales, proventu ad 4 poll, louii^i, cano- 
tomentosi, rbachi crassiuscula ; bractese lanceolatse, 3-1 lin. longge; pedicelli 2 
lio. longi. Calyx sub antbesi tubulosus, vix 6 lin. longus, brevissime et densissime 
incano-tomentosus, ore obliquo, dentibus 4-5 breviter ovatis ; fructifer paulo longior, 
valde ampliatus, subglobosus, deorsum rumpens. Corolla alba, in fauce flavida ; tubus 
cylindricus, 6-7 Un. longus, leviter incurvus, extus tomentosus ; limbi lobi rotundati, 
denticulati, 2J-3 lin. dimetientes. Capsula globosa, 3 lin. dimetiens, stylo coronata. 

In young jungle at the Penokok, alt. 2500 feet {Havlland, 134i). 

B. incana differs from B. dasijantha, Bentli., and B. tomentosa, Benth., ])y the very 
short whitish tomentum and liy the smaller flowers and globose capsules. 



LENTIBULABINE.E. 

Utuicularia ORBicrLATA, Wall. Cat. n. 1500; A. DC. Prodr. viii. 18. 

Maripari Spur, alt. 5500 feet; Temburungo, alt. 7500 feet {Uaviland, 1327, 1328) 

Distribution : Tropical Asia. 

The specimens under no. 1328 correspond with Wallich's type ; those of no. 1327 
have larger flowers, w^ith a much longer spur. But, different as they may appear at 
first si"-ht, they both come wdthin the range of variation which is exhibited, for instance, 
iw the very instructive set of specimens collected by Hooker and Thomson in the Khasia 

iims. 



212 DE. 0. 8T.APF OX THE TLOEA OF MOUNT KIXABALU. 

GESNERACE^. 

JEsCHYNANTHLS MAGNIFICA, Stapf, n. sp. SufFrutex glaberrimus. Eami internodiis 
elougatis 3-pollicaribus, crassiusculi. Eolia petiolo 4-6 lin. longo suffulta, elUptico- 
oblonga, i poll, longa, IJ poll, lata, basi subrotundata, acuminata, coriacea, nervis 
lateralibus utrinque 6 valde indistinctis. Pedunculi axillares, brevissimi, minute 
bracteati, uniflori. Pedicellus 3-5 lin. longus. Calyx obconico-tuhulosus, IJ poll, 
longus, apice 3 lin. latus, 5-dentatus, dentibus 3 lin. lo7igis, obovato-oblongis, 
obtusiusculis. Corolla coccinea, longe exserta, 2J-3 poll, longa, tubus ventricosus et 
siijpra medium lemter curvatus ; lobi rotundati, superiores vix 4 lin,, inferiores 6 lin. 
longi, glanduloso-ciliati. Antlierae 2 lin. longge. 
At 6000 feet (Saviland, 1255). 
Allied to yE. Lcnnponga, Miq., and to jE. pidchra^ G. Don, but very distinct by the 

glabrousness of all j)arts except the margins of the corolla-lobes, the larger size of the 

leaves and JBlowers, the long obconical calyx, the longer calyx-teeth, and the shape of the 

corolla. 

^SCHYNANTHUS TRICOLOR, Hoolv. fil. Bot. Mag. t. 5031. 

On rocks over streams, alt. 3000 feet {Burbidge). 

Distribution : According to a note in Mr. Burbidge's diary, it is common everywhere 
near rivers in North Borneo, either epiphytic on trees or trailing over rocks, 

DiCHEOTEiCHUM ASPEKiFOLirM, Benth. et Hook, fil., Gen. Plant, ii. 101 i ; C. B. Clarke 
in DC. Monogr. v. pars i. 53. 
8000-10,500 feet (Loio ; Ramland, 1076). 
Distribution : Malaya, from Java to Amboina. 

DiCHROTEiCHUM BRACTEATUM, Stapf, n, sp. Caulis rcpens, crassiusculus, radicans, ex- 
siccatus irregulariter sulcatus, apicem versus dense fulvo-tomentosus, infra plus 
minusve glabratus, cortice pallido papyraceo demum soluto. Eolia valde insequalia, 
altera linearia, 3 lin. longa, tomentosa, mox decidua — ideoque caulis specie alterni- 
folius — altera petiolo dense brevissimeque tomentoso, 3-3^ poll, longo suffulta, e 
basi leviter cordata, late ovata, acuta, 3-3^ poll, longa, 2^ poll. Icda, indistincte 
denticulata, crassa, utrinque fulvo-papilloso-tomentosa, nervis lateralibus utrinque 
circa 7. Cymse dicliotomaj, congesta3, in axillis sessiles vel subsessiles, multiflorae ; 
hractew infimcB oblongce vel oblongo-lanceolatcB, I-I5 poll, longw, cum sequentibus 
[interioribus) cito decrescentibus Jlores involucrantes, omnes fulvo-tomentosa? ; pedi- 
celli 3-4 lin. longi, tomentosi. Calyx fere ad basin 5-partitus ; segmenta 4 lin, longa, 
Hneari-oblonga, tomentella. Corolla coccinea (?), 8-9 lin. longa, extus papillosa, 
villosula ; tubus vix exsertus, superne leviter curvatus, latus ; lobus infimus deflexiis, 
3 lin. longus, caeteri porrecti, summi 1| lin. longi, omnes obtusissimi. Stamina 
longe exserta ; antherae f lin. longa?. 

Maripari, alt. 5000 feet (Haviland, 1289). 

Allied to I), asperifolimn, Benth. et Hook, fil, but very distinct by the softer 
indumentum, larger leaves, and the large tomentose bracts. 



DR. 0. STAFF OX THE FLOKA OF MOITNT KIXABALU. 213 

DiDYMOCARPUs AREOLATUs, Stapf, 11. sp. Caulis robustus, dense breviterque ferrugineo- 
tomentosus. Folia opposita, aequalia, petiolo IJ poll, longo breviter tonientoso 
suffulta, oblongo-elliptica, 8 poll, longo, 3 jioll. lata, basi in iictiolum breviter 
decurrentia, apice acuta, argute denticulata, supra uigricantia, in costil nervisque 
adpresse ferrugineo-hirtella, coeterurn dcmuni ghbrata, Inbcrcnlis arcolis poh/gouis 
graciliter circimiscriptis crehre obslta, subtus breviter ferrugineo-tomentella, nervis 
lateralibus utrinque 16-17, graciliter, hinc inde quidem distincte reticulata. Cymae 
in axillis sessiles ad fascicules redact^e, dense ferrugineo-tomentosae ; bracteae 
lineares vel filiformes, 1 lin. longae ; pedicelli 3-4 lin. longi. Calyx ad basin 
h-imrtitm\ segmenta fere fiUfonnla, scepe Jlexuosa, 4-5 lin. louga, tomentosa. 
Corolla alba ; tubus basi angustus, deinde valde ampliatus, 10-12 lin. longus, extus 
fere glaber ; labium superum 4 lin., inferum 6 lin. longum, hocce flavo-maculatum, 
lobis late rotundatis. Stamina fertilia filamentis applanatis geniculatis. nntlieris 
1 lin. longis, dorso glanduloso-pilosulis ; sterilia lateralia circa 4 lin. longa, inter- 
medium duplo brevius. Ovarium basi disco majusculo circumdarum, angnste 
lineare, in stylum minute glandulose pilosulum abiens, cum eo S-10 lin. longum ; 
stigma bilobum. 
Prom 5000 to 7000 feet (Havilaml, 1248). 

The leaves are very remarkable for their regular and graceful reticulation, which is 
finely impressed above, enclosing numerous hexagonal or pentagonal areoles with a 
tubercle in the centre. Underneath it is often inconspicuous on acconnt of the tomentose 
covering, but it is generally very distinct in transjiarent light. The tubercle is the 
widened base of a hair, the upper and thinner part of which shrinks very soon and conies 
off. The corolla is in size and shape not unlike that of Chirita acuminata, K. Br., in 
which direction probably the affinity of the Kinabalu plant lies, though the calyx is 
divided to the very base and the segments are very narrow. 

Ctrtandra Clarkei, Stapf, n. sp. (PI. XVI. c. figs. 17-19.) Caulis e basi repente 
radicante ascendens, densissime fulvo-sericeo-velutinus, inferne plus minusve gla- 
bratus, internodiis 1^ poll, longis. Eolia opposita, sequalia, petiolo 6-8 lin. longo 
velutiiio suffulta, lanceolato-oblonga, l|-2j poll, longa, 8-11 lin. lata, utrinque acuta, 
basi in petiolum plus minusve decurrentia, crenato-denticulata, crassa, supra prime 
densissime flavido-sericeo-villosa, dem.um glabrata, nigricantia, subtus in nervis indu- 
mento simili sed adpressissimo vestita, caeterum pilis adpressis serioeis, demum plus 
minusve evanidis obsita, nervis lateralibus utrinque 6-7. Mores solitarii, axillares ; 
pedicelli ad 6 lin. longi, sericeo-villosi ; bractece Uuxb, oblongte, ad eel nltm medium 
connatce, 7-11 lin. longce, indumento foliorum co?isimiU vestita, cabicis basin inclu- 
dentes. Calyx obconicus, poUicaris, postice plus minusve fissus, legulariter vel 
irreijalariter 5-dentatus, dentibus brevibus triangularibus vel elongatis apice recurvis, 
sericeus. Corollse tubus superne oblique ampliatus, 1 poll, longus, extus sericeus, 
intus albo-maculatus ; lobi coccinei, patentes ; labium suj^erum 7 lin. longum, inferum 
5 lin. longum, lobo intermedio basi cordate, rotundato, lateralibus oblique ovatis. 
Stamina 2; filamenta glabra; antherae ovatae, 1 lin. longae, apiculatse. Ovarium 
glabrum ; stylus minute pilosulus ; stigma dilatatum, medio sulcatum. 

SECOND series. — BOTANY, VOL. IV. 2 F 



214 DE. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

In damp, shady, and steep places, alt. 10,500 feet [Kavlland^ 1131). 

Allied to Cyrtandra rhynchanthera, C. B. Clarke, and to (7. fenestrata, C. B. Clarke, 
both from Sumatra. It is a species very well marked by the large bracts, which enclose 
the flower like another calyx. 

Cyrtandka Btjebidgei, C. B. Clarke in DC. Monogr. v. pars i. p. 262. 

At Kiau, Burhidge\ at 5D00 feet (Hacikmd, 1223). 

Dr. Haviland states that the bracts are white. In his specimen, which is more fully 
developed, the petioles are 5 inches long, and the lamina almost 1 foot by IJ inches. 

Cyrtandra trisepala, C. B. Clarke in DC. Monogr. v. pars i. p. 253. 

Penokok Biver, alt. 3000 feet ; alt. 6000 feet {Haviland, 1215, 1292). 

Distribution : North Borneo. 

The leaves are very variable in size and shape, almost equal or very asymmetric at the 
base, and the smaller ones have their tips spreading or reflexed in most of Dr. Haviland's 
specimens. 

ACANTHACEiE. 

Thunbergia pragrans, Boxbuiigh, Cor. PL i. 67. 
Kadamaian, alt. 2500 feet (Haviland, 1360).. 
Distribution : Tropical Asia and Australia. 

Strobilanthes kinabaluensis, Stapf, n. sp. Caulis herbaceus, tetragonus, fulvo-pubes- 

cens vel tomentellus. Polia opposita, insequalia, majora petiolo |-1 poll, longo 

fulvo-tomentello suffulta, ovata, 3-lJ poll, longa, l|-2^ poll, lata, basi acuta, in- 

sequalia, apice acuminata, inaequaliter crenato-dentata vel repanda, supra fuscescentia, 

adpresse sparsim hirtella, subttis tenuiferfulvo-tomentella, nervis lateralibus utrinque 

6-8, distincte transverse venulosa ; minora l|-lf poll, longa, petiolo breviore, 

caeterum consimilia. Inflorescentise capitatse, ovatse, pauciflorae, muUibracteatce, 

fulvo-tometitellcB, longe vel breviter pedunculata? vel sessiles, in dichasiis perfectis vel 

redactis, vel gemmis accessoriis auctis dispositae ; dichasia solitaria vel 2-3 fasciculata, 

saepe longe pedunculata (ad 2^ poll.), ex axillis superioribus vel pseudo-terminalia, 

interdum in paniculas foliatas collecta. Bmctetje capitulorum esteriores ovato- 

lanc€olat(B, obtusae, ad 8 lin. longas, interiores cito decrescentes, diu persistentes. 

Calycis segmenta lanceolata, 4-5 lin. longa, pubescentia, ciliata. Corolla 10-12 lin. 

longa ; tubus iuferne tenuis,, constrictus, superne subahrupte campanulatim ampliatus, 

subobliquus, extus sparsim pilosulus ; lobi breves, subaequales, late ovati. Pilamenta 

basi pilosula ; antherse 1 lin. longai. Stylus glaber ; stigma arcuatum, unilaterale. 

Capsula \ poll, longa ; semina 4, elliptico-orbicularia, applanata, undique pilis tenuis- 

simis, undulatis,. inelasticis densissime vestita, 1 lin. lonira. 

Kadamaiau Biver,. alt. 2000 feet [Haviland, 13(j2). 

Allied to ♦S'. joaniculata, Miq., from Java, but distinct by much larger bracts and a 
richer indumentum. 



DE. O. STAFF ON THE FLORA OF MOUNT KINA'BALU. 215 

Strobilanthes Galeopsis, Stapf, n. sp .Caulis herbaceus, superne subtetrai^onus, infcrne 
teres, tenuiter fulvo-tomentellus vel demum plus minusve glabratus. Eolia oppositii, 
aequalia, petiolo tenuissime tomentello et prajterea patule sctoso, 1-lJ poll, lonfjo, 
suiTulta, ovata, 2^-3^ imll. longa, lJ-2 j9o//. lata, basi rotimdata vel brevissime 
contracta et in petiohim decurrentia, caudato-acuminata, dentiito-crenata, ciliata, 
supra fuscescentia, adpresse setulosa et prseterea minute adpresseque hirtella, subtus 
birtella et in nervis setulosa, nervis lateralibus utrinque 7-8, venis transversis satis 
conspicuis. Spica terminalis, sessilis, ^-1 poll, longa ; bractca? laneeolato-oblons^ae, 
apice sgepe patulae, utrinque setosoe, diu persistentes ; bracteolas linearcs, Ccilyce ])aulo 
breviores. Calycis segmenta sabtequalia, lanceolato-linearia, liirsuta et setulosa. 
Corolla subpoUicaris ; tubus ad medium angustus, deinde sensim ampliatus et 
leviter cur^atus, extus parce pilosulus ; lobi breves rotundati. Pilamenta longiora, 
pilosula ; antherse IJ lin. longte. Stylus pilosulus. Capsula 4-5 lin. longa ; semina 
4, ovato-orbicularia, fere 1^ lin. longa, fusca, pilis rufis, adpressis, strictis, vix olasticis 
vestita. 
From 5000 to 7000 feet (ffaviland, 1159). 
Allied to S. aspera, Decaisne, from Timor, but distinct by the long-petiolcd, broadly 

ovate leaves, and longer corollas. The general habit is very much that of broad-leaA ed 

forms of Galeopsis Tetrahit or G. versicolor. 



VERBENACE.E. 

Geunsia fahinosa, Blume, Bijdr. 819. 

In young jungle at the Penokok, alt. 2800 feet {Haciland, 1343). 

Distribution : Throughout Malaya, from Singapore and Sumatra to the Philippines and 
the Moluccas, 

Premna caulielora, Stapf, n, sp. Arbor humilis, inflorescentiis exceptis glaberrima, 
undique glandulose punctata. Kami tetragoni, pallidi, inter folia linea transversal! 
notati. Polia simplicia, petiolo \ poll, longo suffulta, late obverse- lanceolata, 
acuminata, callose erenato-denticulata, firma, pallida, nervis lateralibus utrinque 
10-12 prominulis, distincte transverse venosa> subtus tenuiter reticulata. Cymae in 
ramis annotinis vel vetustioribm sessiles, e nodis cra-ssiusculis multiflorce vel pauci- 
florce, |-f p>oll. longce, pube minuta stellata dense adpresseque vestita? ; pedicelli 
demum ad 6 lin. longi, graciles. Calyx cupulari-campanulatus, \-\\ lin. longus, 
subtruncatus. denticulis callosis 4, basin versus parce stellato-puberulus. Corolla 
alba (ex alabastro tantuin nota) ; tubus calycem fere duplo superans ; lobi 4, extus 
glanduloso-leprosi, late ovati, uno extimo, intimo opposito cicteris paulo minore. 
Stamina 4, in alabastro subsequalia ; filamenta glabra ; antherae lineares, basi 
bilobte. Ovarium glandulis aureis depressis dense vestitum ; stylus antheras paulo 
superans, stigma breviter bilobum ; loculum unum ; placentae binse, bifidae, parietales, 
deinde subcontigua3 ; ovula 4, lateraliter affixa. Drupa ovoideo-globosa, interdum 
obtuse tetragona, 2 lin. longa, 4-pyrena, cavitate ceutrali majuscula. 

2e 2 



216 DR. O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

At the Penokok, from 3000 to 3200 feet {Savilcmd, 1301, 1305). 

A very well-marked species, the affi.nity of which I am not able to point out. The 
floral structure is exactly that of a typical 4-merous Premiia, but the corolla is perfectly 
glabrous inside. The inflorescence is said to be terminal in 'Premna, but the fact is that 
it often terminates short branches which come from the axils of leaves and which bear 
only a few leaves or none at all. In this case the inflorescence might be called just as 
well axillary. When these axillary leafless or almost leafless inflorescences appear, the 
supporting leaves are either still persistent or they have already fallen off. Both cases 
may be seen on the same specimen. This kind of development has reached its extreme 
in P. caulijlora. The flower-bearing branches are leafless and reduced to a swelling or a 
knot on the stem, from which the sessile cyme springs. 

Clerodexdron kinabaluense, Stapf, n. sp. Eami novelli dense patule hirtelli, demum 
glabrescentes, cortice pallido. Folia opposita, aequalia, petiolo tenuiter puberulo, 
intcrdura supra basin et sub apice constricto, 1-1^ poll, longo suffulta, obverse lanceo- 
lata vel lanceolato-oblonga, 6-10 poll, longa, 2-2^ poll, lata, acuminata, repando 
serrulata, supra primo setulis minutis, mox — basi incrassata excepta — deciduis con- 
spersa, subtus in nervis fusco-pubera, nervis lateralibus utrinque 11-13. Panicula 
terminalis ad 8 poll, longa, 4-5 poll, lata, erecta, patule pubescens : rami bracteis 
lanceolatis usque filiformibus sustenti, cymas 5-2-floras laxas gerentes ; pedicelli 4-7 
lin. longi. Calyx 5-partitus ; segmenta lanceolata sub anthesi 4--5 lin. longa^ longe 
acuta, pubescentia, demum excrescentia. Corollae tubus tenuis, 6-8 tin. longus, 
pilosulus; lobi subsequilongi, subspathulati, apiculati. Stamina stylusque | poll, 
exserta. 
At the Penokok, alt. 3200 feet {Uavilcmd, 1307). 

Allied to C. cUsparifolium, Blume, C. Grlffithiafium, C. B. Clarke, and C. calamitosum, 
but distinct by the long leaves ; from the first also by the much larger calyx, and from 
the second by the shorter corolla-tube. C. obtusidens, Miq., of which I do not know the 
type, has the calyx only half as long, according to the description. There are in the 
Ilerljarium several closely -allied species from North Borneo, bnt all undescribed. 



LABIATE. 

Phlomis urGOSA, Benth. in Wall. PI. As. Ear. i. 63. 

Prom 3000 to 5000 feet {Baviland, 1228). 

Distribution : Prom Sikkim and the Khasia Hills to the Mishmi Hills and North 
Burma ; Malaya, from the Peninsula to the Philippines. 



POLYGONACEiE. 
Polygonum chinense, Linne, Sp. PI. i. 520 ; Meissn, in DC. Prodr. xiv. i. 130 
In young jungle on the Tampassuk River, alt. 2500 feet {Ilaviland, 1350). 
Distribution: Tropical Asia ; China; Japan. 



DE. O. STAPP OX THE FLORA OF MOUNT KINABALU. 217 

NEPENTHACE^. 

Nepenthes Lowii, Hook. fil. in Trans. Linn. Soc. xxii. 420, t. 71. 

From 6000 to 8000 feet (Low) ; epiphytic on Dacrydium, Rliododendron, and other 
scrubby, moss-draped trees, down to 5000 feet {Burbidge). 



Nepenthes villosa, Hooker fil. in Hook. Ic. PI. t. 

Prom 7000 to 9000 feet {Low) ; with N. Majah in yellow loam and rough sandstone 
grit {Burbidge). 

Nepenthes Edwardsiana, Low, ex Hook. fil. in Trans. Linn. Soc. xxii. 4-20. 

Erom 8000 to 9000 feet {Low) ', epiphytic on low mossy shrubs and trees at 6000 feet 
{Hurbidge). 

Nepenthes Rajah, Hook. fil. in Trans. Linn. Soc. xxii. 421, t. 72. 

At 5000 feet {Loio) ; on open patches among sedges and other herbaceous plants, in 
yellow loam or in gritty soil, from 5000 to 9000 feet {Burbidge), 

Nepenthes phyllamphora, Willd. Spec. PI. iv. pt. ii. 874.— Syn. N. Burbidgei, Hooker 
fil. ex Burbidge in Gard. Chron. xvii. (1882), 56. 

At 3500 feet {Low, Burbidge). 

Distribution: Malaya, from Mt. Ophir and Sumatra to the Louisiades and to South. 
Cliina. 

Nepenthes tentaculata, Hooker fil. in DC. Prodr, xv^ii. 101. 
Erom 6000 to 8000 feet {Burbidge) \J at 9000 feet {Low). 
Distribution : North Borneo. 

PIPEBACE.E. 

Peperomia CANDIDA, Miq. Syst. Pip. 105. 
At 6000 feet {miviland, 1220). 
Distribution : Malaya, from Sumatra to Borneo. 

CHLORANTHACE.E. 

Chloranthus brachystachytjs, Blume, El. Jav. viii. 13, 14, t. 2. 

At 6000 feet ? [Raviland, 1205 ?). 

I bave some doubt about this label belonging to this plant, as Haviland has written 
on it : — " Small tree — flowers white." 

MONIMIACE^. 

Scyphostegia, Stapf, gen. nor. 
Elores unisexuales. Elos 6 ignotus. Elos ? : Perianthiuni fere ad basin 6-lobum ; 
lobi 2-seriatim imbricati, persistentes. Discus receptaculum urceolatum, carnosum, 
ad faucem incrassatum et recurvum formans. Carpella cc in fundo receptaculi, 
basi squamis hyalinis plerumque 3, inferne varie coalitis, circumdata ; stigma obli- 
quum, depressum; ovulum erectum, anatropum. Eeceptaculum friwtifermn 



218 DE. 0. STx\PF ON THE FLOEA OF MOUNT KINABALU. 

auctura, globosum, a fauce I'umpens. Carpella stipitata, cylindrica, leviter curvata, 
erecta, sicca ; pericarpium chartaceum. Semen erectum, testa tenuissima, 
albumine carnoso ; embryo axilis medio semine longior, cotyledonibus obovato- 
oblongis, radiculse inferse sequilongis. Prutex scandens. Folia alternantia, 
cbartaceo-membranacea, crenulata. Mores bracteati, in racemis longis, laxis 
dispositi ; racemi inferne compositi, ramis flores 2 valde approximatos gerentibus, 
quorum alter plerumque abortivus ; bractese infundibuliformes, truncatse. 

ScYPHOSTEGiA BORNEENsis, n. sp. (PL XVII. 1-14). Prutex glaberrimus. Ramuli 
graciles, obscure angulosi. Polia oblonga, ad 5 poll, longa, ad 1| poll, lata, sub- 
abrupte breviterque acuminata, l)asi rotundata, crenulata, exsiccando fuscescentia, 
lucida, nervis lateralibus utrinque 6-9, subtus prominulis, venis trans versalibus 
crebris, tenuibus. Eacemi ad 1 ped. loDgi, subflexuosi, ad ramorum basin decidue 
minuteque bracteati ; rhachis angulosa ; rami graciles, 1-lf poll, longi. Bracteae 
flores subtendentes infundibuliformes, 2 lin. longge, truncatae. Pedicelli breviter 
vel vix e bracteis exserti. Periantbium in alabastro ovatum apiculatum, sub anthesi 
patens, segmentis exterioribus quam interiores subduplo majoribus, obovato-oblongis, 
2-3 lin. longis. Peceptaculum globosum, sub anthesi 2 lin. altum, margine recurvo, 
incrassato, lobulato coronatum ; fructifermn 9 lin. dimetiens. Carpella pilosa, ad 
4 lin. longa. 
Near Koung, alt. 2000 feet {Raciland, 1377). 

The floral structure of this plant is in many ways so remarkable that I feel justified 
in founding upon it a new genus, though the male flowers are unknown. There is no 
genus in Monimiaceae to which it might be said to be manifestly allied. Yet the 
analogies with genera of this order are so obvious that I consider it will most appropri- 
ately find its place amongst them. There is very little in the vegetative parts and in 
the inflorescence to guide us as to the systematic position, though the leaves are not un- 
like i\\o%Q oiDaphnandra repandula, P. v M., in texture, venation, and general appearance. 
The flowers on the lower branches of the inflorescence stand usually in pairs, very close 
to each other, and supported by a pair of obconical truncate bracts which fit exactly 
into each other, and more or less enclose the pedicels. It seems that only one of these 
flowers develops normally, whilst the other remains in the state of a bud. In the upper 
part of the inflorescence no second flower is formed; yet we may assume per analogiam 
that these solitary flowers are in fact reduced inflorescences. These two-flowered and 
one-flowered partial inflorescences are then collected into a long and very loose raceme. 
There is nothing very peculiar in the perianth; the "disc," however, deserves full 
attention. It is a fleshy, urn-shaped receptacle, the margin of which is very much 
thickened and recurved, leaving open only a narrow passage. Prom the bottom of this 
receptacle the carpels rise, filling scarcely more than one half of the cavity. They seem 
to be enveloped in a young state by a delicate membrane, which later splits more or less 
irregularly into 2-3 lobes, reminding one of the perianth of a fig-flower. The lower 
part and even the lobes themselves are more or less connate, and it is difficult, if not 
impossible, to separate them without tearing them. The carpels are distinctly stalked. 



DR. O. STAFF OX THE FLORA OF MOUNT KIXABALU. 219 

The stigma is oblique, and there is generally a more or less marked one-sided constriction 
below it. It is depressed in the middle, and the depression leading to the stiguiatic 
channel is clothed with minute papillse. The ripe fruit consists of the enlarged 
receptacle, supported by the persistent perianth. It seems to break up from the mouth, 
thus exposing or freeing the enclosed carpels. Judging from the ripo seed, the solitary 
ovule is erect and anatropous. 

The arrangement of the apocarpous ovaries in an urn-sliaped cavity is very character- 
istic of various Monimiacese ; but in these the cavity is formed by the convex torus in 
connection with the perianth-tube, and the perianth-lobes surround the mouth of this 
receptacle. In some species, however, tlie margin rouiul the mouth is distinctly raised 
into a rim. If we imagine this rim enlarged and transformed into an urn-shaped 
receptacle and at the same time the torus I'aised to the level of tlic insertion of the 
petals, we should get the same structure as we find in Scyjihostegia. As to the carpels 
themselves and the seeds, there is nothing in them which would disagree with the 
general characters of Monimiacese, though the embryo is remarkably large for the order. 
The involucre at the base of the carpels is, however, very peculiar. The materLil in my 
hands is hardly sufficient to clear up their development so thoroughly as^ I wislied, 
particularly with regard to the earlier stages. Yet I have the impression that this 
involucre is not formed by true phyllomes, but is rather of the nature of trichoines and 
emergences. In various Monimiacea^ the tissue of the torus protrudes between the 
carpels so as to envelop them more or less completely, in others^it sends fortli hairs 
between them. The involucre in Scyp1iodegla\'& therefore possibly an analogous growth. 

I have examined the anatomy of the branches and of the leaves, and Dr. Solereder 
of Munich was so kind as to do the same. But neither of us was able to find anything 
characteristic in the anatomical structure; the secretion-cells present in the leaves 
of most Monimiacea? were not met with in Scyphostegia. 

LAURACE^E. 

LiTS.EA CAULIFLORA, Stapf, n. s-p. (§, Cylicodapline) 6 . Rami noveUi brevissime- fulvo- 
tomentelli. Folia alterna, petiolo crasso 6-8 lin. longo sufi"ulta, oblongo-lanceolata, 
7_9 jyoii, longa, 2-3 polL lata, basi breviter attenuata, apice subacuminata, tenuiter 
coriacea, supra opaca,. demum glabrata velia costa diutius adpresse fulvo-tomentella, 
subtiis purpureo-glaucescentla, undlque tenuiter pubescentia, nervis lateralibus 
utrinque 18-20, subrectis, delude sub margine subito prorsus arcuatis, venis- transver- 
salibus distinctis, tenuiter reticulata, nervis venistjue supra impressis^ subtus 
prominulis. Umbellulae iti racemis densls, valcle abbreciatis, m acervis interdnm 
1 ptoll. dimetientibits, e ramis vetustis ortls conglomerate ; pedunculi demum 
2-3 lin. longi, ut bracteoe involucrantes fulvo-sericei ; flores circiter 6, brevissime 
pedicellati. Perigonium extus fulvo-sericeum, 6-partitum; segmenta obverse 
lanceolata, 2 lin. longa. Stamina 12, interiora sensim minora ; filamenta extimi 
cycli 3^ lin., intimi 1^ lin. longa, omnia pilosa, cyclorum interiorum basi biglandu- 
laria, antherse omnes 4-locellat9e. Ovarium rudimentarium. 



220 DE. O. STAPP ON THE FLOEA OF MOUNT KINABALU. 

At 5000 feet {mwiland, 1221). 

The foliage is not unlike that of LitscEci khasiana^ Meissn., or L. elongata^ Wall., 
with which the affinity probably lies. 

LiTS^A ZEYLA.NTCA, C. & Pr. Nees in Anioen. Bot. Bonn. fasc. i. 58, t. 5, ex Nees Syst. 
Laur. 626.— J , $ . 

At 8000 feet {Haviland, 1128). 

Distribution : Throughout Malaya, from Sumatra to the Philippines and the Moluccas ; 
on the mainland to the Khasia Hills and Bhotan ; Ceylon and in the Deccan Peninsula 
as far as the Concan. 

I take L. cinnamomea, Blume, from Ternate, and a specimen from Luzon {Tidal, 528), 
to be the same species. 

LiKDERA (?) GEANDiFOLiA, Stapf, n. sp. $ . Rami crassi, novelli brevissime fulvo- 
tomentelli. Polia opposita vel subopposita, petiolo crasso 3-4 lin. longo suffulta, 
obovato-oblonga, 9-13 jioll. longa, 3^-5 poll, lata, basl breviter acutata, apice 
breviter et latiuscule acuminata, tenuiter coriacea, supra costa nervisque puberulis 
exceptis glabrata, fuscescentia, subtus purpureo-glaucescentia, sparse et minutissime 
pilosula, nervis lateralibus utrinque 10-12, angulo 45° ortis, venis transversalibus 
laxis distinctis, utrinque tenuiter reticulata, nervis supra leviter impressis, venis 
venulisque utrinque prominulis. Umbellula? in racemis valde abbreviatis, multi- 
floris, den sis vel depauperatis, solitariis vel pluribus dense glomeratis, circiter ^-florae ; 
pedunculi 1-1| lin. longi, ut bractese involucrantes fulvo-sericei ; pedicelli subnulli. 
Perigonium extus sericeum ; segmenta 6 ingequalia, 3 plerumque multo minora^ 
majora ovata acuta, 1 lin. longa. Staminodia 12 filiformia, interiora brev iora et 
basi glandulis binis instructa. Ovarium glabrum, cum stylo l^j lin. longum ; stigma 
dilatatum, sublobatum. Bacca (immatura) perigonii tubo truncato, explanato, vix 
aucto insidens, oblonga, 4 lin. longa. 
Penokok Ptiver, alt. 3000 feet {Hamlcmd, 1334). 
This is possibly a Litscea, but the fruit resembles closely that of Lindera reticulata. 

LiNDERA c^siA, Eciuw. cx ViUar in Blanco, PL Eilipp. ed. III. App. 181. — Syn. 
I)aj)hnidium ccEsium, Xees in Miq. PI. Ind. Bat. i. 976. 

, var. EUEA, Stapf. Piamuli novelli et petioli dense rufo-strigillosi. Polia majora ad 

\\ poll, longa, acuminata, sed non adeo ut in var. acuminata, Miq. 1. c. 
At 8000 feet {Haviland, 1106). 
Distribution (of species) : Malaya, from Sumatra to the Philippines. 

PROTEACE^. 
Helicia erratica, Hook. fil. in El. Brit. Ind. v. 189. 

At 5000 feet {Haviland, 1197). 

Distribution : Sikkim Himalaya, Khasia Hills and Martaban, East to Burma ; Borneo. 

Dr. Haviland states that the flowers are white, whilst those of the Sikkim and Khasia 
plants are said to be pale yellowish. The material is rather imperfect, and the Bornean 
plant may prove to be a new species or variety. 



DE. 0. STAFF O^S THE FLOEA OF MOUNT KIXABALU. 221 

THYMEL^ACEiK. 

Drapetes ericoides, Hooker fil. in Hook.. Ic. PI. 895. 
From 12,000 feet to the top (Zoic ; Ilaclland, 1078). 
Distribution : North Borneo ; Xew Guinea, Mt. Knutsford. 

LORANTHACE.E. 

LORANTHUS ESTIPITATUS, Stapf, n. sp. (§ Deudrophthoe).~Yv\xlQK cpipliyticus. Hamuli 
novelli ferrugineo-furfuracei, mox glabrati, vetustiores crebre lenticeJlis alliidis 
conspersi. Eolia opposita, j^etiolo circiter 3 liu. loni^o suffulta, oblonga \c\ clliptiea 
vel plus minusve formam obovatani vel ovatam appetentia, f-2 poll, longa, ^-IJ 
poll, lata, basi rotuudata vel plus minusve breviter cuneata, apice obtusa, coriacea, 
juvenilia ferrugineo- vel fulvo-furfuracea, mox glabrata, nervis lateralibus utriiupu; 3, 
rarius 4, plerumque valde indistiiictis. Cymula? axillares, pedunculo 1-2 liu. lougo 
(rarius longiore) suffulta?, furfuracete vel glabratae; pedicelli 2-4 lin. longi, sub 
tlore ipso bracteam minutam ovatam vel rotundatam, margine ciliolatam gerentes. 
Calycis limbus brevissimus, ciliolatus. Corolla in alabastro tubo incurve vel 
subrecto, gracili, basi haud nisi anthesi imminente ampliato, apico obtuse capitato- 
clavata, plus minusve furfuracea, aperta, G-10 lin. longa, postice ultra mediuui 
fissa; lobi 4, lineari-spatliulati, a medio refracti. Stamina lobis corollinis breviora; 
antherse lineares, 1 lin. longse. Ovarium bremter ohlougmn vel (jlobosmn, vix 1 lin. 
longum, primo f urfuraceum, demum glabrescens ; stylus corolla paulo longior ; 
stigma capitatum. Bacca ovoidea vel oblonga vel subglobosa^ ad 3 liu. longa, 
glabrata. 
On the Tampassuk River, alt. 2500 feet (Havilcmd^ 1351). 

Distribution : Erom Tonkin and South China to Borneo and the Philippines ; Perak. 

I refer the following specimens to this species : — Tonkin : Tu-Phat [Balansa, 2329], 

Ououbi \_Balansa, 1030] ; South China : Lo Pan Shan, Prov. of Canton [Hort. Bot. 

Hongk. a. 1883, 7], Hainan [Senry, 8261] ; North Borneo : Kudat \_I)i\ Fraser, 183] ; 

Philippines \_Ciiming, 1970]; Perak [Scorfeohhii, 311, 421]. 

The Philippine specimen is enumerated as L. sphenoldeus, Blume, in Vidal, PI. Vase. 
Eilip. 231, together with no. 490 of Cumiog; but only the latter can be referred to 
L. spJienoideus. Scortechini's plant, on the other hand, was named Z. Scurrula var. 
obtecta by Dr. King, Journ. As. Soc Beng. Ivi. p. 92. But, as in Cuming's specimen, 
no. 1970, the fruit is oblong and not at all clavate. I have carefully gone through our 
whole material of L. Scurrula, which comprises several bundles, and found the shape 
of the fruit very uniformly clavate; and whilst I should not object to merging 
L. repandus, obovatus, and sp^henoideus (all of Blume) into L, Scurrula, I think the 
species just described is distinct enough to deserve specific rank, the chief differential 
character being the shape of the ovary and fruit. It is not impossible that it will prove 
to be identical with Deiidrophthoe fuscata, Miq. (= Lorauthus fuscatus, Korth.), from 
Sumatra. The fruit of this species is said to be a " bacca ovato-globosa," but Miquel 

SECOND SERIES. — BOTAKY, VOL. IV. - G 



222 DE. O. STAPE OX THE ¥LOEA OF MOUNT KINABALU. 

adds "inferne attenuata." I have not seen the tyjje of it, and I do not consider the 
description to be sufficient for identification. The specimens on which Lor ant Jim 
estipifatns is founded differ to a certain degree in the intensity of the tomentum, the 
size of the leaves, and the length of the corolla, but not so much as the various forms 
comprised under L. ^currula do. The fruit is more ovoid in the Bornean specimen, 
and more oblong, or almost cylindric, in that from Perak ; but these extreme forms are 
perfectly connected by intermediate states, and they are always sessile, not stipitate. 

liORANTHUS SABAENSis, Stapf, n. sp. (§ Lepeostegeres). Erutex parasiticus, glaberrimus. 
Eami crassi, teretes vel irregulariter angulati. Eolia opposita vel subopposita, peti- 
ole i-51in.longo crassosuffulta, ovata, 2 iwll. longa,\-H poll, lata, basi rotundata, 
rarius subcordata, apice acuta, coriacea, costa subtus prominente, ctetcrum enervia, 
evenia. Capitula strobiliformia, axillaria, ovata vel oblonga, tetragona, coroUis 
omissis i-j poll, longa, 3-4-fl.ora. Bracteas exteriores involucraates 8-10, interiores 
accrescentes, rotundato-ovatse usque oblonga^, obtusae vel subacutse, pergamaceae, 
castanese, dorso carinatae, intimae f3ores subtendentes, lineares, circa 6 lin. longtie. 
Calycis limbus vix ^ bn. altus, subscariosus, irregulariter lobulatus. Corolla flavida, 
in alabastro clavata, aperta, tubo subrecto vix 1 poll, longo, gracili, sub fauce 
leviter dilatata ; i^egmenta 6, basi ovata, ad 2-3 lin. erecta, deinde linear ia et rejlexa, 
expansa 8-9 lin. longa. Stamina 6 ; filamenta linearia, porrecta, cum antheris 
linearibus 2 lin. longa, corollae segmentis expansis paulo breviora. Ovarium 
turbinatum, 1 lin. longum; stylus filiformis, IJ-lf poll, longus; stigma capitatum. 
Erom 9000 to 11,000 feet on Rhododendron ericoides and Leptospermuni recurvum 
Haviland, 1079, 1108). 
Distribution : North Borneo (Sarawak, Becc. 2579). 

A very iiicomj)lete young specimen, collected by Low at 8000 feet on Kinabalu, evi- 
dently belongs also to L. Kavilandii. 

Allied to a Lepeostegeres from Malacca [Ilaingay^ 695], which, according to Dr. King, 
is perhaps identical with X. Klngii, Scort. ex King, in Journ. As. Soc. Bengal, Ivi. p. 99. 
Maingay's specimen differs by its larger leaves and much larger flowers (to 2 J in.), which 
are split for about 4-6 lin. into 5 linear segments, then widened for 1-1^ in. and very 
narrowly cylindric at the base. I have not seen the type of L. Kingii, but from the 
description it appears to me rather doubtful whether Maingay's specimen is really iden- 
tical with the true L. Kingii. However that may be, Haviland's plant certainly differs 
also from the latter by the smaller leaves and flowers and the sexfid corolla, which is not 
slit on one side. 

LoRANTHUS CENTiFLORUs, Stapf, n. sp. (§ Lepeostegeres). (PL XVIII. c. 9-10.) Prutex 
parasiticus, glaberrimus. Rami stricti, teretes, nigricantes. Folia subopposita, 
petiolo crassiusculo 6 lin. longo suffulta, ovato-oblonga, 3-3^ poll. lon"-a, 1^-1^ poll, 
lata, basi rotundata, apice obtusa, coriacea, supra fusco-nigricantia, subtus rubro- 
fusca, nervis lateralibus utrinque 6-8, tenuissimis. Flores mimerosissimi, ad 50, 
in capitnUs fere 2 poll, dimetientibus subterminalibus congesti, branteis amplis 
involaruatis. Bracte^ chartacea), circiter 6, exteriores breviores, rotundatse, transverse 



DE. O. STAPF ON THE FLOKA OF MOUNT KINABALU. 2'2Z 

latiores, interiores longiores, rotundo-obovatae usque spatliulata), ad 1 poll, longia, in 
margine scariosae et interdum uiidulatie. Receptaculuiii crassum, J poll, dimetiens. 
Pedicelli nulli vel brevissimi, crassi, angulati, rarius ad 2 liu. longi, apice hracteolaw 
scepissime ad Uneam elemtam, ovarii basin ciugentemvcl unllatcraUter snOiemlcjilew, 
redactam gerentes. Calycis limbus ad 1 lin. altus, 5-dentatus. CoroUce tubus 2-3 
lin. longus\ segmenta fiUformia, ad 8-1) lin. porrecta, abliinc refracta, parte refraeta 
3 lin. longa anguste lineari. Stamina 5 ; fiJamcnta segmentls adnata, ad h Hn. 
libera ; antherne subulatse, 2 lin. longie, plus minusve torttc. Ovarium angulatuin, 
1^ lin. longum, mox basi dilatatum et superne constrictum. Eacca (iinniatura) 
ampuUacea, plus minusve angulata, 3-1 lin. longa. 
At 6500 feet {Ravlland, 1217). 

A very fine and remarkable species, the affinity of whicli is not quite evident at first 
sight. The inflorescence is not unlike that of the Lepeostegeres group, and it comes 
particularly near to that of L. Beccaril, King. But the bracts are much loss numerous, 
and the heads are short and broad, containing about 50 flowers each. They spring from 
the uppermost axils of young branches, thus appearing almosl terminal. The ovaries 
are, as in L. Beccaril, closely packed together and more or less angular from the mutual 
pressure, and there are no pairs of bracteoles at the base of the pedicels, which are usually 
reduced to a minimum and grow only partly longer after fertilization has taken place. 
The corolla is split for f of the entire length, and there is a sudden bend, about \ below 
the apices of the segments, by which the upper part is perfectly defloxed, so that it 
appears almost as an appendix to the corolla-segments. The filaments arc not free from 
the base of the segments as in typical Lepeodegeres, but only from a little below the 
bend just mentioned. This is also the case in L. sabaensis, where, however, the corolla- 
tube is much longer. The disappearance of the bracteoles may be explained as suppression 
caused by the close packing of the numerous flowers, whilst the coalescence of the 
filaments and the corolla for the greatest part of their length might be considered as 
having remained unaltered in spite of the far advanced j^artition of the corolla. Thustlie 
species may be characterized as a Lepeostegeres with very numerous and closely packed 
flowers, with consequently-suppressed bracteoles, and with a very deeply-slit corolla. 

BALANOPHOBACE^. 

BAlanophora elongata, Blume, Enum. PL Jav. i. 87. 

At 11,000 feet [Uavlland, 1166). 

Distribution : Java ; Borneo. 
Balanophora Lowii, Hooker fil. in Trans. Linn. Soc. xxii. 426, t. 75 c. 

In shady woods {Low). 

EUPHORBIACE^. 

Glochidion tenuistylum, Stapf, n. sp. 2 . Frutex humilis. Ramuli tenues vel validi- 
ores, interdum leviter anfracti, dense fulvo- vel rufo-hirtelli. Folia petiolo hirtello 
1-11 lin. longo suff"ulta, oblonga vel ovato-oblonga, 3|-1 poll, longa, l^-lj poll, lata, 
basi rotundata, sfepe paulo insequalia, breviter acuteque acuminata, firme membra- 

2&2 



22i -DE. 0. STAFF OX THE FLORA OF MOUXT KINABALU. 

nacea, pallida, supra (demum quidera), costa nervisque hirtellis exceptis, glabra, 
minute pustulata, subtus in nervis veuisque hirtella, cseterum pubescentia, nervis 
lateralibus iitrinque circiter 7 cum venis transversalibus arcuatis vel flexuosis laxe 
anastomosantibus, supra saepe impressis, subtus valde prominulis. Elores feminei 
3-5 in fasciculis axillaribus dense glomerati. Calyx yix 1 lin. longus, dense liir- 
tellus ; segmenta 6, lineari-oblonga. Ovarium densissinie liirtellum ; styli 4, raro 5, 
ad medium in columnam e basi quam ovarium vlx tenuiore sensim pauloque incrassatam 
vel siibcylindricam connati, ahhinc filiformes, subdivergentes, minute hirtelli, IJ-lf 
lin. longi. 
At tbe Penokok, alt. 4000 feet (Haviland, 1303). 
Yery similar in habit, foliage, and tomentum to GlocJiidion Moonii^ Thw., from Ceylon, 

but distinguished by the slender styles, which are connate to one half of their length 

only, and decidedly shorter. 

Daphniphylltjm borneense, Stapf, n. sp. Frutex vel arbor humilis, glaberrimus. 
E-amuli crassiusculi, circa 2 lin. dimetientes, cortice nigricante, inferne lenticellis, 
interdum crebris, pallidis obsiti. Polia ad apices innovationum valde approxin'.ata, 
petiolo 6-8 lin. longo suffulta, elUptica vel obovaio-elliptica, 2^-3^ poll, longa, 1-lf 
poll, latn, basi breviter acuta, apice abrupte breviterque apiculata, tenuiter coriacea, 
exsiccando nigricantia, subtus plus minusve glaucescentia, nervis lateralil)U8 utrinque 
circiter 7, angulo circiter 60° ortis, insigniter laxissime reticulata^ nervis venisque 
supra impressis vel prominulis, subtus prominentibus. Plores masculi subumbellati 
vel brevissime racemosi, pedunculo J-1 poll, longo suffulti ; pedicelli 2-3 lin. longi. 
Sepala ovata, vix ^lin. longa. Stamina 5-9 ; filamenta J lin. longa ; antherge paulo a 
dorso compress^e, lineares, 1 Un. longce, minute apiculata?. Flores feminei in race- 
mis demum elongatis, pedunculo 4-6 lin. longo suffulti; pedicelli ad 3 lin. longi. 
Sepala ovata, ^ lin. longa, mox decidua. Ovarium ovatum, 1 lin. longum ; stylus 
crassiusculus, brevis ; stigmata recurva, ovariofere cequilonga. Pructus ellipsoideo- 
globosus, 4 lin. longus, basi rotundatus, apice stigmatibus coronatus. 
Prom 10,000 to 11,000 feet (Eaviland, 1070). 

I have much hesitation in founding a new species on this plant, as it comes so near to 
D. himalayense, Muell., and particularly D. glancescens, Blume. But, after repeated 
comparison with the very rich material of these species in the Kew Herbarium, I prefer 
to keep it separate. It differs from D, glaucescens by the longer and narrower anthers, 
which are at least twice as long as broad, and by the much longer stigmata, and perhaps 
also by the very coarse and prominent reticulation ; from J), himalayense by the broad 
leaves, the less numerous nerves, and the longer stigmata. I find the anthers and the 
stigmata extremely uniform in I), glaucescens, though tbe species covers a very lar^-e area, 
and this was the reason which chiefly induced me to separate the Bornean plant as a 
species. 

Baccaxjuea lanceolata, Muell. Arg. in DC. Prodr. xv. pt. ii. 457. 

At the Penokok, alt. 3000 feet [Haviland, 1340). 
Distribution : Malaya, from Sumatra to Borneo. 



BE. O. STAFF ON THE FLORA OF MOUNT KTXAB.ALU. 225 

Antidesma aueitum, Tulasne, in Ann. Sc. Nat. 1851, p. 203 d (ex descr.}. 

At the Penokok, alt. 1000 feet [nacllatid, 1337). 

Distribution : Java. 

I have not seen any specimen of (his species, but the plants in my hands agnjc so per- 
fectly with Tulasne's explicit description that I have little doubt in referring them to this 
species. The leaves vary in Haviland's specimens from 7 to 9 in. in length, and from IJ to 
2J in. in breadth. 

Antidesma Moritzii, "Muell. Arg. in Linnaea, xxxiv. 67. 

At Koung, 2000 feet [Hamland, 1374). 

Distribution : Malaya ; on the mainland as far as Tenasserim ; in the Archipelago from 
Sumatra to the Philippines. 

Beccari, Sarawak, no. 3146, and a specimen by Barber, probably from Labuan, belong 
also to this species. 

Claoxylon pauciplokum, Stapf, n. sp. {^Eu-Claoxylon. — 6 2). Ramuh' pallidi, novelli 
dense et adpressissime subsericeo-pilosi, mox glabrati. Folia alterna, jo<?//o/o i-f lln. 
longo, supra canaliculato suffulta, oblongo- vel obovato-lauceolata, 1-7 j)(>ll. ioiiga, 
1^-2^ poll, lata, basi acuta, biglandulosa,apicc breviter caudato-acuminata, in inargine 
calloso-denticiilata, pallide viridia, supra primo adpresse pubescentia, demum glabrata, 
subtus, pra^cipue in nervis venisque, minute adpresseque pilosa, ncr\ is lateralibus 
utrinque 6 erectis, venis transversalibus tenuibus, distinctis, subhorizontalibus. Mas : 
Spicse (juveniles) ^-^ poll, longce, adpresse pilosa? ; bracteas, ut videtur, ad normam 
unifiorce vel hijtorce (?), late ovatae, IJ lin. longae, sua longitudine distantes ; bracteolae 
multo minores. Calyx ante anthesin ovato-globosus, adpresse pilosus, vix 1 lin. 
longus, 3-partitus, valvaris. Receptaculum in tota superficie inter stamina glanduMs 
ovatis vel breviter cylindricis, apice pilis 1-3 duplo triplove longioribus instrnctis 
obsitum. Stamina numerosissima ; antherae loculis inter se liberis, basi fixas. 
Femina : Spicae valde redactcB, rhachi cij'citer 2 lln. longd, hi- vel iuiiflor(B ; 
bractese late ovatae, obtusae, 1-lJ lin. longse. Calyx 3-partitus ; segmenta ovata, 
1-1^ lin. longa, adpresse pilosula. Ovarium disco crassiusculo irregulariter lobulato 
cinctum, globosum, 2 lin. dimetiens, 3-loculare, villosum ; stigmata simplicia, 
subsessilia, intus grosse papillosa ; ovula in quoque ioculo solitaria. 
At the Penokok, alt. 3000 feet ( 2 ) {Havilaud, 1317) ; at 6000 feet ( j ) (Eaviland, 

1311). 

In habit and foliage not unlike C. longifolium, Muell., but very distinct ]jy the short 

few-flowered inflorescences, and also by the shorter petioles. 

Acalypha Caturus, Blume, Bijdr. 629 ; Mueller Arg. in DC. Prodr xv. pt. ii. 805. 

Penokok, alt. 3000 feet {Saviland, 1312). 

Distribution : Malaya, from Sumatra to Borneo ; probably also in New Guinea. 

This seems to be a very variable plant. Haviland's specimen has very large leaves (the 
largest 7 by 5 J in.) and extremely slender and long spikes (to 14 in.). A specimen obtained 
by Beccari (Plant. Sum. 754) from the province of Padang belongs to it, and probably also 
a specimen collected ])y Forbes (127) iu the Sogeri region. Eastern New Guinea. 



226 DE. O. STAFF ON THE FLOKA OF MOUNT KINABALU. 

AcALYPHA STiPULACEA, Klotzscli, in NoY. Act. Ac. Nat. Cur. xix. Suppl. i. 416. 
In young jungle on the Tampassuk EjiYC-r, alt. 2500 feet {Haviland, 1363). 
Distribution : Malaya, from JaYa to the Philippines and Moluccas. 
This plant was also collected by Dr. M. Eraser at Banggi, North Borneo (no. 232). 

UETICACEJ::. 

Picus SETiFLORUS, Stapf, u. sp. {§ Ett-Syce ?) (PI. XVIII. b. 5-8.) Prutex hurnilis. Ramuli 
graciles, novelli fulYO-strigillosi, Yetustiores cortice pallide brunneo. Polia petlolo 
fidvo-strigilloso f-1^ ^;oZ/. lougo suff'tilta, OYata, 3-4J poll, longa, lf-2J poll, lata 
basi rotundata, apice acuminata, margine argute serrato-dentata.^rmwZe membranacea, 
supra sparse strigulosa, aspera, subtus in nerYis Yenisque hispidula, nerYis lateralibus 
utrinque 4-6, pare intimo fere e basi orto, Ycnis transYersalibus tenuibus remoti- 
usculis, reticulatione tenuissima. Receptacula feminea axillaria, subsessilia, iuYolucro 
triphyllo cincta, phyllis minutis late rotundatis, matura, globosa, 3-3 J lin. dimetientm, 
flaYida, adpresse sparseqiie hispidula^ intus flavido-setosa. Perianthium pedicellatum, 
segmenta 3-4, oblonga Ye'i oboYata, tenuissima, irregulariter crenulata ; dorso, imprimis 
apicem versus, setulis rigidis ^ lin. long is, segmoitis ipsis longioribiis, ohsita. 
AcliEenium subglobosum ; stylus lateralis, ^ lin. vel paulo longior. 
At 7600 feet (Haviland, 1214). 

Evidently allied to F. chrysocarpa, Heinw., with which it has the bristly perianth in 
common, Lut the branches are much more slender, the petioles much longer, the leaves 
thinner, and the receptacles smaller and not tomentose. The perianth is comparatively 
larger. On the other hand, it resembles also certain species of the section Sycidium in 
its general appearance, and it might, perhaps, be better placed in it. 

Ficus EiivABALUENSis, Stapf, n. sp. (§ Eu'Syce.) Erutex glaberrimus. Hamuli nigro- 
cinerascentes. Eolia sessilia, ohovata, l-|-2 poll, longa, 8-9 lin. lata, basi subcuneata, 
apicn ohtusa vel apiculata, margine incrassato, angustissimc revoluto, coriacea, 
exsiccando supra nigrescentia, subtus f usca, nervis lateralibus utrinque 8-10 patulis, 
supra impressis, subtus prominentibus, tenuioribus, paucis ssepe interjectis, subter 
marginem distincte arcuatim anastomosantibus. Beceptacula mascula bina in 
axillis f oliorum, pedunculo 1-2 lin. longo suffulta vel subsessilia, involucro triphyllo 
basi cincta, phyllis rotundatis 1 lin. longis ciliolatis, globoso-ovata, 3 lin. longa. 
Elores masculi sub receptaculi ore dispositi. Perianthium glabrum ; segmenta 3-4, 
obovata ; stamina 3. Elores gallipari numerosi ; perianthii segmenta 3, obovata vel 
posterius plus minusve navicularia ; ovarium subglobosum ; stylus brevis lateralis. 

At 8000 feet {E:avila7id, 1216). 

Allied to F. vaiHolosciy LindL, but differing in the broad, usually very obtuse and 
sessile leaves. 

Eicts MoTLETANA, Miqucl, in Ann. Mus. Lugd. Bat. iii. 228, 294. 
At 8800 feet {Haviland, 1215). 
Distribution : All over Borneo. 



TfR. O STAFF OX THE FLORA OF 510UNT KINABALU. 227 

Laportea stimulans, Miquel, ia Zollinger, Cat. 103. 

At 5000 feet, 3 4 feet high {Havlhaid, 1222, 1230). 

Distribution : Malaya, from the Peninsula to the Philippines. 

Dr. Haviland's specimens differ from the type specimens in their leaves being thinner 
and acute at the base or almost decurrent, but specimens from the Malayan Peninsula, 
chiefly from Perak, vary so much in these points that I doubt whether tliese characters 
can be relied upon. No. 1222 agrees exactly with a plant collected by Motley near 
Banjermassin (no. 346). Weddell says, in a note on the sheet, "forsan ad L. crenidaiam, 
var. (3. referenda." But I am inclined to refer this var. /3., i. e. I nzouensis, to L. stimii- 
lans rather than to Z. crenata, Gaudich. However that may be, L. cvemdata and 
L, stimulans are extremely close. 

Lapoktea, sp. aff. L stimulanti^ Miq. 6 . 

Penokok lliver, alt. 3000 feet [Ilaviland, 1341). 

This is probably a new species, very distinct by the sessile or subsessile leaves, which 
are obovate-lanceolate and long cuneate at the base. The material, however, is too 
incomplete for descri2ition. 

Pilea Johniana, Stapf, n. sp. (PI. XIX. d. 17-26.) lEerba percnnis, nana, ca^spitosa, 
e basi ramosa ; rami prostrati radicantes, tenues. Folia densa, superne swpe densls' 
sima et internodlis multo longiora, petiolo \ lin. longo vel brevissimo latiusculo 
sutfulta, orbiculai'ia vel deltoideo-orbicularia, 1-li Hn. longa, basi brevissime 
latissimeque cuneata, apice obtusissima, carnosa, exsiccando nigricantia, triplinervia, 
nervis inconspicuis utrinque, sed pra^cipuc subtus, cystolithis linearibus minutis 
conspersa. Stipulje minutse, lanceolatae, J lin. longae. Flores masculi in axillis 
soUtarii, suhsessiles. Perianthium bipavtitum vel 1-partitum ; segmenta rotundata, 
cucullata, apiculata ; stamina 2 vel 4. Plores feminei in cymulis axillaril)us folio 
multo brevioribus, ^«?f<?//7or?s ; pedicelli \ lin. longi vel breviores. Perianthium 
tripartitum ; segmentum intermedium vix ad \ lin. longum, ovatum, cucullatum, 
quam lateralia, plerumque in margine glandulosa, 2-3plo longius. Staminum rudi- 
menta minutissima. Ovarium ovatum; stigma breviter penicillatum, sessile, 
xlchenium suboblique globoso-ovatum, J- J I'm. longum, Iseve. 
In crevices of rocks from 11,000 to 12,000 feet {SacUand, 1391). 
A very marked species, allied to P peploides, Hook, et Arn. It differs by the very 

crowded, almost sessile leaves, which are thick and not transparent in a dry state, the 

few-flowered inflorescence, and the larger achenes. I saw but one male flower, from 

which the description is drawn. 

Pilea PTEROCArLis, Stapf, n. sp. Herba glaberrima, 3 ped. alta. Caulis crassus, 
quadrmlatus, alis circa f lin. latis. Folia opposita, sequalia vel subaequalia, petiolo 
2-3 lin. longo, basi 1^-2^ lin. lato suffulta, lanceolata vel oblonga, 5^-7^ poll, longa, 
li poll, lata, basi minute biauriculata, apice acuminata, acumine minute denticulato. 
cseterum integerrima, membranacea, supra nigro-viridia, subtus pallidiora, utrinque 



228 DE. 0. STAPE ON THE FLOE A OF MOUNT KIN AB ALU. 

cystolitMs crebris linearibus conspersa, trinervia, nervis ad acumen productis, venis 
transvcrsalibus distinctis. Stipulae inter folia cicatricem transversam conspicuam 
relinquentes. Mores mascidl in glomerulis parvis in spica simplici vel parce 
composita, axillari, gracillima, IJ poll, longa dispositi. Perigonium 4-partitum ; 
segmenta ovata, f lin. longa. Stamina 4, Ovarium rudimentarium minutum. Mores 
feminei in cyniis axillaribus brevissimis, multifloris, glomeruliformibus. Periantliium 
3-partitum ; segmenta valde ina^qualia, intermedium demurn ^ lin. longum, lateralibus 
3-plo majus. Aclisenium suboblique globoso-ovoideum, valde compressum, f lin. 
longum. 

On the Tampassuk River, alt. 3000 feet {Saviland, 1229), 

Allied to P. smilacifolia^ Wedd., but very distinct by the winged stem and the very 
short and broad petioles. 

PiLEA CRASSiroLiA, Stapf, n. sp. Herba glaberrima. Caiilis quadrangularis, sulcatus, 
flavido-virescens. Polia opposita, subsequalia, petiolo IJ-l^ poll, longo snffulta, 
oblonga vel oblanceolata, 2-3 poll, longa, 7-9 lin. lata, basi minute biauriculata, 
apice acuminata, in margine, basi excepta, crenato-denticulata, crassa, exsiccando 
nigro-viridia, subtus pallidiora, utrinque cystolithis creherrimis ohsita, et minute ni- 
gropunctcda. 'FlovGSiuascidi in glomerulis densis, parvis, in spicis plerumque parce 
compositis, solitariis vel paucis, in axillis foliorum dispositis. Perigonium 1-partitum; 
segmenta ovata. Stamina 4. Ovarium rudimentariuir minutum. Plores feminei 
pauci cnm masculis in centro glomerulorum. Perianthium tripartitum; segmentum 
intermedium lateralibus majus, cucuUato-obovatum, apiculatum. Ovarium 
ovoideum ; stigma incrassatum. 
At 3000 feet {Hamland, 1339). 

Allied to P. smilacifoUa, Wedd., but distinct by the very thick and erenulate-denti- 
culate . leaves, the transversal veins of which are hardly visible, and the abundance of 
cystoliths, amongst which black dots are scattered over the lower surface of the leaves. 
A Philippine specimen from the island of Panay ( Vidal, 3863) comes still nearer, but it 
has rather asymmetric leaves, with very unequal and often short petioles. 

Elatostemma lineaee, Stapf, n. sp. 6 . Caulis, ut videtur, arborum truncis ramisve 
appressus et applanatuSifulvo-strigillosiis, internodiis 4-6 lin. longis. Polia alterna, 
disticha, petiolo 1-2 lin. longo snffulta, late Unearia vel lance olato-linearia, 1^-2| 
poll, longa, 4 lin. lata, basi insequalia, in latere extus spectante in petiolum auricu- 
latum decurrentia, acuta, a medio vel in summa tertia ^wnta grosse arguteque deniato- 
serrata, crassiuscula, supra demum quidem glaberrima, luride glauco-viridia, sparse 
nigro-punctata, subtus in costa necnon in nervis adpresse strigillosa, pallidc fusca, 
punctulis crebris obsita, nervis lateralibus subpatuUs utrinque 8-11, scepe valde 
indistinctis vel parenchymate plane obtectis. Stipulse oblique ovatae, acutse vei 
acuminatse, membranacese, 3-4 lin. longaj. Cymj3e capitatim contractie, brevissime 
pedunculatse vel sessiles, l|-2 lin. dimetientes; bractese extimaj late ovatae, apice 
incrassatse, pilosula?, vix 2 lin. longse lataeque, interiores oblanceolatse vel spathulatae, 



DR. 0. STAFF OX THE FLORA OF MOUXT KTXABALU. 220 

angustse, ciliatse, vix breviores. Periantliium 4-5-partitum, 1 lin. longum ; segmenta 
late obovata, paulo cucullata et sub apice cornuta, clorso pilosa. Ovarii rudimentum 
minutum. 
Kinitaki River, alt. 5000 feet {Eaviland, 1271). 

Allied to E. rupestre, Wedd., but differing by the evidently flattened and strigillose 
stem, the narrow leaves, wliicli have almost parallel margins, a very coarse serrature, 
and a usually very indistinct nervation. 

Elatostemma. Lowii, Stapf, n. sp. — $ . Herba erecta. inflorescentiis exceptis glaherrima. 
Caulis inferne crassiusculus, nigricans, superne (exsiccatus quidem) sulcatus, ramis 
gracilibus. Folia alterna, sessilia vel petiolo brevissimo tenui suffulta, <)bli(|He 
oblongo-lanceolata, l|-2 poll, longa, 5-7 lin. lata, basi in latere exteriore subuuri- 
culatim prodticta, in latere interiore plerumque sensim attennata, apice brctiter 
vel caudato-acuminatay a medio vel fere a basi grosse crenato-dentata, tenuiter 
membranacea, supra nigricantia, subtus fuscescentia, ssepe ob stomata prominula 
minute albo-punctulata, cystolithis utrinque {specie quidem) nnllis, trijilinervia, 
nervis lateralibus patulis utrinque 5-7, tenuibus. Stipula; sul)ulatae, ^-1 lin. longae. 
Cymae axillares, foliis ssppe admodum diminutis sustenttc, sessiles, in receptaculiim 
confluentes, 1-1^ lin. dimetientes ; bractese omnes anguste lineares vel spathulato- 
lineares, moUiter pilosse, 1-^ lin. longse ; pedicelli vix f lin. longi. Perianthium 
tripartitum ; segmenta filiformi-linearia, pilosula. Staminodia paulo longiora vel 
breviora. Achenium ovatum, J lin. longum, pericarpio tenui, leviter tuberculato. 
At 5000 feet, on the northern face of Mount Kiuabalu (Low). 

1 bave not been able to identify this plant with any of the species represented in the 
Kew Herbarium ; only tw^o, viz. E. JJrmlleanum, Brongn., and E. lancifolium, Wedd., 
seem to come very near it. The former is figured in Brongniart, Bot. Voy. Coq. t. 46, 
tig. A, and is evidently very like in habit. But the leaves are more cuneate at the 
base in E. Urvilleanum, less acuminate, and full of cystoliths. The figure of the 
female flower as given by Brongniart is possibly wrong, as it shows only two periantii- 
segments and no staminodes. B. lancifoUimi, on i\\e other hand, has a strigose stem 
and entire or sub-entire leaves, which are hairy below. 

Elatostemma thalicteoides, Stapf, n. sp. — 6. (PL XIX. b. 5-8.) Caulis gracilis, 
firmus, teres, adpresse hirtellus ; ramuU gracillimi, distlche foliati. Polia alterna, 
brevissime petiolata vel sessilia, suboblique cimeato-ovata, 6-9 lin. longa, 4 lin. lata, 
supra medium in tmo latere extus spectante dentibus 2 majusculis, in altero ple- 
rumque dente unico magno vel alio parvo addito, dentibus mucronulatis, membra- 
nacea, supra la3te viridia, pilis perpaucis adpressis conspersa, ca^terum glaberrima, 
subtus pallidiora, in nervis et in margine adpresse tenuiterque strigillosa, nervis 
lateralibus tenuibus numero dentium. Stipulae lanceolat^e, strigillosae, vix ^ lin. 
long«. Cymae axillares, pauciflora?, in glomerulos 1-1^ lin. dimetientes congestae, 
dense pilosae ; bracteae extimae rotundato-ovatae vel obovatse, sub apice breviter mu- 
cronulatte, dorso dense et rigide pilosse ; interiores spathulatae, basi valde attenuatae, 

SECOND SERIES. — BOTANY, VOL. IV. 2 H 



230 DE. O. STAFF OX THE FLORA OF MOUNT KINABALU. 

|-| lin. longse ; peclicelli |-J lin. longi. Perianthium ad medium 5-lobum ; lobi 2 
exteriores ovati, acuti, interiores cucullati, dorso sub apice filiformiter cornuti, 
omnes extus rigide pilosi. 
At 5000 feet (Haviland, 1219). 

A very graceful plant, the foliage of which imitates the compound leaves of certain 
forms of Thalictrum minus, particularly in the shape of the leaflets. It is allied to JE. podo- 
phyllum^ Wedd., but distinct by the cuneate base of the leaves and the absence of 
conspicuous cystolitlis. 

Elatostemma bulbotheix, Stapf, n. sp. (PL XIX. a. 1-4.) Herba humilis, ut videtur 
prostrata. Caulis firmulus, teres, superne dense strigilloso-hirtellus, internodiis 2-3 
lin. longis. Polia alterna, disticha, sessilia vel subsessilia, oblique obovata, 6-8 lin, 
longa, 3-5 lin. lata, basi in altero latere extus spectante rotundato-auriculata, in altero 
acutata, obtusa, obtusiuscule dentato-crenata, firma^ crassa, supra flavido-viridia, 
rugosa, densissmie cystolithis obducta, s'parse nigro-punctata, glabra, subtus in costa 
nervisque et in margine adpresse strigillosa, in parenchyniate pilis basi incrassatis 
et nitide aureis cceterum alhis minute Mspida, triplinervia, nervis venisque supra 
valde impressis, subtus eximie prominentibus. Stipulse minutae, iusequales, una 
extus spectans subulata, altera lanceolata vel ovata, strigillosse, vix 1 lin. longee, 
Cymee biflorse, sessiles, axillares ; bractese exteriores anguste oblongoe, vix 1 lin. 
longse, interiores diminutee, spathulato-lineares vel fere subulatse, omnes firmulae, 
breviter strigillosge ; pedicelli brevissimi. Perianthium nullum. Staminodia 4, sub- 
quadrata. Ovarium ellipsoideum, minute quadricostatum, la^ve, vix ^ lin. longum. 
At 10,000 feet {Haviland, 1392). 

A very marked species, the affinity of which is difficult to ascertain. Perhaps it lies 
with ^. glaucescens, Wedd. I do not know any other species where the cystolitlis are 
so crowded on the upper surface of the leaves, which they cover entirely, with the sole 
exception of the black dots. The stiff sliort bristles of the lower surface are thickened 
at the base, and, in a young state at least, shining and yellow, so as to make it glitter 
from minute golden-yellow dots. 

Elatostemma lithoneurum, Stapf, n, sp. (PL XIX. c. 9-16.) Herba, ut videtur, 
prostrata. Caulis superne patule hirsutus, inferne glabratus, nigrescens, et albido- 
lerrucosus. Polia alterna, disticha, sessilia vel subsessilia, oblique oblanceolata, 
iuterdum subfalcata, 2-2^ poU. longa, J-} poll, lata, basi in uno latere extus 
spectante auriculato-producta, in altero acuta, breviter vel indistincte acuminata, 
fere a basi argute dentata, crassa, supra nigricantia, adpresse setulosa, subtus 
pallidiora in costa nervisque necnon in parenchymate setoso-hirsiita, ceeterum 
glandulis 3-4-cellidaribus fuscis conspersa, triplinervia, nervis lateralibus (prajter 
subbasalia) 5-7 in latere majore, valde prorsus arcuatis, nervis supra subimpressis, 
subtus prominulis et cystolithis crassis albis obsitis. Stipulse oblanceolatee, saepe 
subfalcatae, obtusiusculse, 5 lin. longse, membranacea^. Cymse masciilce in capitulis 
depresso-globosis 6 lin. dimetientibus, pedunculo J-l| poll, longo hirsuto suti'ultis ; 
bractese extimt3e late rotundato-ovatae, obtusae, sub apice breviter obtuseque cornutge, 



DE. 0. STAPF ON THE FLOEA OF MOUNT KINABALCJ. 231 

nervis 3-5 siibprominentihiis, parce setulosaB, i-ij liu. lonp^jie lataeque, nigrescentes, 
interiores gradatim angustiores et in formam spatliulatam ciicullatain abeinitos, 
omnes cystolithis albis paucis vel pluribus conspersic et margiiio ciliatula^ ; pedieclli 
breves. Perianthium 4-partitum, 1 lin. lons^um, extus parce setulosiim ; segmenta 
obovata, valde cuciillata, bina majora sub apice cornuta, bina aiigustiora gibboso- 
cornuta. Cymae feminece axillares, in receptaculum collectte, sessiles, 2 lin. 
dimetientes, miiltiflorae ; bractetc extimoe ellipticju, sub apice cornuta?, 1] lin. 
longae, interiores sensim angustiores et in formam cucuUato-spatbulatam abeuntes, 
dorse setulosae; pedicelli demum elongati, \ lin. longi. Periantbium nullum. 
Staminodia 3 minuta. Achenium ellipsoideum, supra medium tenuiter 4-costu- 
latuni, ^ lin. longum. 
Bebind boulders of torrents, alt. 11.000 feet (Ilanland, 1200). 

In the shape of the leaves and in the floral structure some resemblance to J^. popil- 
losa, Wedd., is evident, but E. lithoneurum differs by the mucli firmer leaves and the 
coarser indumentum. There are no cystoliths visible, at least from outside, except on 
the nerves beneath, which are covered with thick, wedge-sbaped cystoliths. In any case 
it is a very well-marked species. 

MYRICACE^. 

Myrica Javanica, Blume, Bijdr. 517, Fl. Jav. (Myric.) p. 1. t. 7, 6 [forma alpina). 

Prom 11,000 to 11,500 feet (Loic ; Ilaviland, 1091). 

Distribution : Malaya ; Java, Borneo. 

The leaves are much smaller than in the type, but all other characters point so clearly 
to M. javanica that I have no doubt in referring this plant to it. It is evidently a 
stunted alpine form. The leaves are l-l:j by ^-f in. Low's specimen, consisting of a 
few male scraps, is enumerated erroneously under Quercm montlcola, King, in Ann. Bot. 
Gard. Calc. ii. 41. 

CUPULIFER^. 

QuERCUS Havilandii, Stapf, n. sp. (PI. XVIII. a. 1-1.) Arbor humilis. Eamuli 
dense brevissimeque fulvo-furfuraceo-tomentelli. Polia petiolo crassiusculo, furfu- 
raceo-tomentello, 2-2^ lin. lougo ^ViEwiiSi, late ovata, 1^-2^ poll, longa. l^-llj poll, 
lata, basi late rotundata vel subtruncata, abrupte breviterque acuminata, coriacea, 
supra exsiccando castanea, primo fulvo-f urfuracea, mox, costa nervisque exceptis, 
glabrata, subnitida, subtus densissime atque persistenter ochraceo-furfuraceo-tomen- 
tosa, nervis lateralibus utrinque 10-12, supra impressis, infra prominentibus, venis 
transversis tenuibus. Stipulse minutse deciduae. Spica? androgynae vel subuni- 
sexuales vel unisexuales, terminales et ex axillis summis, simjjlices vel compositae, 
folia paulo superantes, dense furfuraceo-tomentellae. Flores mascnli solitarii (an 
semper ?) in axillis bractearum minutarmnillismulto breviorum ; bracteolae 2 minutse. 
Perianthium 6-partitum; segmenta ovata, obtusa, i lin. longa. Stamina 8-10. 
^love^ feminel in spicis propriis quam masculse brevioribus vel ad basin spicarum 
androgynarum ; bracteae minutse, ovatge. Perianthium 1-6-lobatum. Styli, apice 

2h2 



232 DR. O. STAFF ON THE FLOEA OF MOUNT KTNABALU. 

excepto, pubescentes. Cupulse per paria a dorso connatse (vix niaturse), 6 lin. dime- 
tlentes et 2 lin. altce, fiilvo-tomentosse, ob apices squamamm liberos nigro-tuberculatfe ; 
glans depresso-hemisphserica, apice excepto glabra, lucida, apiculata, 3-4 lin. alta. 
From 9,000 to 10,500 feet (Haviland, 1069). 

Closely allied to Quercus pruinosa, Blume. I should have considered this oak as a form 
of Q. pruinosa, but for the very minute bracts, the solitary male flowers, and the small 
leaves and fruits, and I am still very doubtful whether all these characters may prove to 
be the result of the physical conditions of a much elevated and exposed locality. 

A specimen from Lepanto, N. Luzon ( Vidal, 1814), belongs evidently to the same species, 
but the leaves are less rounded at the base. There are no iiovvers or fruits with it. 

Castanopsis sp. aff. C. argentecB, A.DO. 

" Small tree, fruit prickly like an opened chestnut," at 8000 feet {Haviland, 1115). 

I have seen only a barren branch. The nerves are rather faint and more curved than 
in C. argentea. It is very probably a new species. 

Castanopsis TURBiTSTATA, Stapf , n. sp. Arbor humilis. Hamuli robusti, minute denseque 

toraentelli. Eolia petiolo crasso 3 lin. longo suffulta, ovata vel ovato-oblonga, 3-4 

poll, longa, 1^-1 j poll, lata, basi rotundata vel breviter acuta, apice sensim tenuiter- 

que acuminata, valde corlacea^ supra exsiccando fusca vel obscure olivacea, subnitida, 

siib lente valido pilis stellatis minutissimis alhis parce conspersa, subtus viridi-cinerea, 

pube stellata laxa minutissima vestita, nervis lateralibus utrinque 9-10 sub margine 

insigniter arcuatim anastomosantibus, supra graciliter impressis, subtus prominenti- 

bus, venis transversis subtus tenuibus distinctis. Elores feminei in apice ramulorum, 

. in spicis valde abbreviatis, minute cinereo-tomentellis, subglomerati. Cupula 

turhinata, bracteis squamiformibus, ovatis, superioribus plerumque breviter et abrupte 

vel sensim acuminatis, exlmie in spirocyclis primo arete congesiis, demum plus 

minusve solntis dispositis ornata, minutissime tonientella, et insuper glandulis minimis 

aspersa. Perianthium 6-lobum, ad loborum apices in cupula inclusum. Styli basi 

pubescentes. 

At 8500 feet {Haviland, 1102). 

Allied to C. stmiatrana, A.DC, but distinct by the minute stellate tomentum, the 

very coriaceous leaves and the turbinate cupula, which bears a spiral of bracts of 5-7 

whorls. The largest cupula is about 7 lin. long and 4 lin. broad in the upper part ; 

when yoimg the cupula is more cylindric, measuring 2J by 2 lin. 

BUHMANNIACEiE. 
BrRMANNiA PAPiLLOSA, Stapf, n. sp. Herba annua, chlorophyllo destitute, aphylla. 
Caulis circa 3-pollicaris, simplex vel furcatus, basin versus squamis paucis 
minutis obsitus, uniflorus, sub flore bibracteatus vel nudus ; bractese lanceolatae, 
1^ lin. longge, hyaliuse. Perigonii tubus 2 lin. longus, ovarium fere duplo superans, 
trialatus, alls apice abrupte basin versus longe attenuatis, ad jlin. latis ; lobi exteriores 
rotundati, obtusi, in margine et in dorso superne minutissime papillosi, f lin. longi 
latique, interiores 6-plo breviores, late rotundati, minutissime papillosi. Antheree 
sessiles, transverse latio7^es ; connectivum apice in cristulas 2 ovatas vel rotundas 



DE. 0. STAPE 0!^ THE FLOEA OF MOUNT KIXABALU. 233 

fimbriato-denticulatas productum, basi hi-evisslme calcaratum. Ovarium 1 liii. 
longum ; stylus cylindricus, apice breviter tribracliiatus, labio stigmatis infero intus 
papilloso. 
Penokok River, alt. 3200 feet {Hamland, 1329). 

This species is evidently closely allied to Burmannia lutescens, Becc, and B. tridentata, 
Beec. It differs, however, from the former in the shape of the flower, which is obovate, and 
in the anthers, which are broader than long ; from the latter by the presence of distinct 
though very small inner perianth-lobes and the exceedingly short spur of the anthers. 

Burmannia longifolia, Beccari, Mai. i. 21J!, t. xiii. f. 1-5. 
North-west face of Mount Kinabalu at 5000 feet (Loic). 
Distribution : Throughout Malaya, from Perak to North Borneo and New Guinea. 

OBCHIDE^. 

(By n. N. lliDLEY, F.L.S.) 

MicuosTYLis Calophylla, Beiclib. fil. in Gard. Chron. xii. (1879), 718. 
Penokok Iliver, alt. 3000 feet (Ilavllarid). 

Platyclinis grandiflora, E/idl.,n. sp. Bhizoma lignosum, fibrosura, radicibus crassis. 
Pseudobulbi congesti, subteretes, 1^ poll, longi. Folium 4 poll, longum, | poll, latum, 
lanceolatum, acutura, nervis 4 cum costa prominula. Scapus gracilis, nutans, circiter 
8 poll, longus. Plores subremoti, circiter 20, mediocres. Bracte* |- poll, longie, 
convolutse, late oblongae, acutae, ovariis cum pedicellis paulo breviores. Sepala 
lanceolata, acuta, f poll, longa, lateralia obliqua. Petala angusiiora et breviora, 
lanceolata, acuta. Labellum quadratum, angulis acutis, apiculatum, apiculo quarei 
anguli producti vix longiore, lamellis semicircularibus, crassis ad laminae basin, et 
inter eas callo mammillari minus elevato instructum. Columna arcuata, processubus 
linearibus e basi exortis. 

At 10,500 feet {Havilmid). 

The flowers, which are apparently brown, are rather large and few for the genus. 

Platyclinis corrugata, Bidl., n. sp. PJiizoma gracile, pseudobulbis ovoideo-globosis, 
flavis, semi-pollicaribus, corrugatis dease tectum. Folium lineari-lanceolatum, sub- 
acutum, petiolatum, lamina 2 poll, longa, I poll, lata, petiolo pollicari, nervis in laminie 
dorso tribus elevatis. Scapus gracilis, 8 poll, longus, racemo subdenso, multifloro. 
Flores parvi. Sepala ovata, caudata, § poll, longa. Petala lanceolata, acuminata, mi- 
nora. Labellum cymbiforme, breve, tenue, si explanatum ovatum, acuminatum, lobis 
lateralibus brevibus truncatis, et cum margine lobi medii denticulato. Callus in disco 
carnosus, magnus, hippocrepiformis, paulo elevatus. Columna parva, brachiis parvis 
oblongis truncatis e basi exortis, clinandrii margine brevi, integi-o, vix elevato. Rostel- 
lum elongatum, suberectum, integrum, linguiforme. Columnae venter infra stigma 
protrusum. 
Maripari Spur, alt. 5500 feet {Haviland). 



234 DE. O. STAFF OX THE FLOEA OF MOUNT KINABALU. 

Platyclinis stachyodes, Hidl., n. sp. Rhizoma lignosum, pseuclobiilbis cylindrico- 
conicis, arete appressis, 1 poll, longis, J poll, crassis (exsiceatis). Folia linearia falcata, 
coriacea, S-nervia, striolata, 3 poll, longa, I poll. lata. Scapus graeilis, teres, 5-poll. 
longus, superne racemosus. Racemus compactus, 2 poll, longus, miiltiflorus. Brac- 
tesd glumacese, ovatse, T-neryiae, aciitse, -g^ poll, longae, ovaria siiperantes. Plores 
minores. Sepala lanceolata, acuta, ^ poll longa. Petala subsimilia, latiora. Labellum 
subsequilongum, lanceolatum, acutum, lobis lateralibus brevibus, subovatis, falcatis, 
obtiisis, irregulariter siniioso-dentatis, lobo medio multo longiore, lanceolato, acuto ; 
callus ad basin cuneato-quadratus, emarginatus, nervis elevatis 2 ex angulis in lami 
nam extensis. Columna brevis, lata, margine clinandrii dorsali quadrato, elevato, 
tridentato, brachiis lateralibus nuUis. Rostellum semiovatum, crassum. 
At 11,000 feet {Ecwiland). 
An elegant little plant, with the large conspicuous bracts so common in the genus, and 

a compact raceme, reminding one of an ear of wheat, whence the specific name. 

Pendeobiijm kinabalijense, Ridl., n. sp. Rhizoma longe repens, ramosum, angu- 

latum, gracile, radicibus paucis, teretibus, gracilibus, longis. Pseudobulbi ^ poll. 

longi, quadrati, \ poll, lati, 1 poll, dissiti. Eolia solitaria, elliptica vel elliptico-ovata, 

brevissime petiolata, obtusa, coriacea (sicca polita, striata), 2 poll, longa, fere 1 poll. 

lata, vel minora. Racemus ex apice pseudobulbi exortus, brevis, ^ poll, longus. 

Plores majusculi 2-3. Bractese late lanceolatae, \ poll, longse. Pedicelli f poll, longi, 

graciles. Sepala lanceolata acuta, J poll, longa. Mentum breve, latum, J poll. 

longum. Petala lanceolata, mucronata, sepalis minora. Labellum trilobum ; lobi 

laterales lati, J poll, longi, subfalcati, obtusi, lobus medius angustior, paulo longior ; 

carina mediana paulo elevata, obtusa, abrupte in dimidio disci terminata, lamellis 

brevibus duabus semicircularibus versus basin lobi medii. Columna lono^iuscula, 

tilaraento Ion go, clinandrio prof undo, ventre excavate. Capsula angusta, falcata, 

IJ poll, longa, costis sequalibus linearibus. 

At 7000 feet ; epiphytic [JECaviland). 

This is quite an aberrant plant, with a branching stem, from w^hich on one side rise 
a number of distant short branches which represent the pseudobulbs, from the apex of 
which arise the racemes. I think it is best referred to the section Sarcopodlmn at 
present, but it differs much in habit from the others of this section. 

BuLBOPHYLLTJM MONTENSE, Ridl., n. sp. Rliizoma elongatum, gracile, tenue, internodiis 
semipollicaribus oblongis, apicibus sursum curvatis. Polium in unoquoque inter- 
nodio oblanceolatum, obtusum, petiolatum, apice emarginatum, ^-1 poll, longum, 
^ poll, latum. Scapi capillares, 1^-2 poll, longi, e basibus internodiorum assurgentes. 
Plores solitarii, terminales, pedicellis pollicaribus, gracilibus. Ovaria perparva. 
Sepalum postieum J poll, longum, ^ poll, latum, lateralia paulo longiora et latiora, 
oblongo lanceolata, acuta, 5-nervia, liava. Petala oblonga, obtusa, tenuia, vix J poll, 
longa. Labellum linguiforme, carnosum, curvum, breve, exsiccando purpureum. 
Columna pede elongate assurgente. Stelidia longa, subulata, acuta. Capsula (im- 
matura) i poll, longa, obliqua. 



DE. 0. STAFF OX THE FLORA OF MOUNT KINABALU. 235 

Epiphytic ; sepals and petals veined orange yellow; at 11,000 feet {ITavihnid, 1099). 
This species with the next, belongs to a small series of very small plants cliaracterized 
by the moniliform rhizome, every node of which normally produces a single leaf, and at 
intervals throws up a slender pedicel terminated by a single flower, the lateral sepals of 
which are longer than the npper one. 

Bulhophyllum striatellum, Ridl., belonging to the same section, has been unfortunately 
somewhat confused with B. cojwhinmn, Hook, fil., Ic. PI. t. 2038, as by on accident tlie 
stelidia in the diagram of its fertilization (Annals of Botany, iv. no. xv. pi, xxii.) have been 
represented too long. The shape of the petals alone would show that the two plants are 
utterly distinct. B, concinnuni, Hook., belongs to the capitate group with distinct 
pseudobulbs {B. leptanthum. Hook., &c.). B. striatellum resembles this group in its 
vegetative organs, but is distinct in its single flower with oblong caudate sepals. 

BuLBOPHYLLUM CATEXARiUM, Eidl., n. sp. Rhizoma ramosum, internodils oblongis, 
apicibus sursum curvatis, moniliforme, pallide viride. Folia erecta, -} poll, longa, 
\ poll, lata, elliptica, obtusa, pallide virentia. Scapus filiformis, poUicaris, uniflorus. 
Bractea minuta ovata. Pedicellus longus capillaris. Flos parvus, late expansus. 
Sepalum posticum lanceolaturn, obtusum, lateralia latiora, triente longiora, auran- 
tiaca, rubro-striata. Petala lanceolata, trientem sepali postici icquantia, flava, linea 
mediana rubrtt. Labellum breve, linguiforme, ovatuni, acutum, carnosum, papillo- 
sum, medio canalicnlatum, atro-kermesinum, canaliculo obscnriore. Columna recta, 
longiuscula, pede longo, apice sursum curvato, libero, roseo. Anthera depressa, 
margine integro, papillosa. Stelidia longa, snberecta, dentiformia, acuminata, acuta. 
Amongst wet mosses on shrubs and trees, alt. 6500 feet {Haviland, 1164). 
Distribution : Malay Peninsula, on Mt. Bukit Hi tarn in the district oi Selangor, at 
2000 feet {Kelsall) ; in mangrove jungle at Kranji near Singapore. 

This curious little species seems to have a wide distribution. I am quite unable to 
distinguish between the Kinabalu plant and that of the Singapore mangrove swam23s. 
It is very abundant in the latter locahty, but seldom flowers. 

BuLBOPHYLLUM CORIACEUM, Ridl., n. sp. Rhizoma crassuni, hgnosum, radicibus 
crassis. Pseudobiilbi nuUi. Folia 1^ poll, dissita, coriacea, crassa, oblongo- 
lanceolata, 2^ poll, longa, f poll, lata, petiolis -| poll, longis canaliculatis. ^Qi\\n 
foliis approximati, 5 poll, longi, validi, basi vaginati, superne laxe racemosi. Flores 
flavi, circiter 13, pro sectione magni. Bractese lanceolata?, acuta?, I poll. longa% 
quam pedicelli longiores. Sepalum posticum lanceolaturn, acutum, lateralia sul)- 
similia, inferne prolongata, J poll, longa. Petala circiter dimidium sepali aequantia, 
oblonga, uninervia. Labellum linguiforme, carnosum, obtusum, marginibus incras- 
satis, canaliculo mediano, baud stipitatum. Columna pede recto, stelidiis triangu- 
laribus lanceolatis, acuminatis. 

At 10,500 feet {Haviland, 1100). 

A stout but not very large plant, with larger flowers than in most species of the section. 

BULBOPHYLLUM MONTiGENUM, Ridl., n. sp. Rhizoma crassum. Caules basi dilatati, 
subbulbosi, breves, vaginis tecti, approximati. Fohum oblongum, lanceolaturn, 



236 DE. 0. STAFF OX THE FLORA OF MOUXT KINABALU. 

obtusum, petiolatum, 7 poll, longiim, f poll, latum, petiolo 1 poll, longo. Scapiis folio 
approximato subseqiiilongus, gracilis, basi vaginis acuminatis jjluribus fere poUicaribus, 
superne laxe racemosus. Tlores plures, dissiti, J poll, longi, pedicellis brevissimis. 
Bractese paleaceae, pedicellis longiores. Sepala lanceolata, caudata, lateralia mentum 
saccatum formantia. Petala carnosa, papillosa, sepalis multo breviora. Labelluni 
petalis a^quilongum, oblongum, lanceolatum, obtusum, pubescens, margines basin 
versus involuti, approximati. Columna minima, stelidiis dentiformibus. 
At 6000 feet (Hamland, 1252). 

In tbis species the short stems on which the leaves are borne are sliglitly dilated, but 
bardly form real pseudobulbs. 

BuLBOPHYLLiiM BEEViFLOEUM, Eidl., n. sp. Khizoma baud crassum, radicibus copiosis. 
Folia approximata, ebulbosa, lineari-lanceolata, petiolata, obtusa, carinata, lamina 4 
poll, longa, f poll, lata, petiolo bipollicari. Scapi debiles, pedales, graciles, a basi ad 
dimidium compressi, vaginis apice acutis exceptis nudi, superne laxe racemosi. 
Plores dissiti, parvi, caduci. Bractese lanceolatse, acutse, vix ^-g poll. longfE. Pedi- 
celli ^ poll, longi. Sepala lateralia ovata, acuta, ;J poll, longa, posticum lanceolatum, 
ovatum. Petala linearia, oblonga, dimidio sepalorum longiora. Labelluni tenue, 
oblongum, obtusum, petalis sequilongum, nervis tribus incrassatis, brevissime 
stipitatum. Columna alis latis, apice stelidiis erectis, majusculis, lineari-lanceolatis. 
Anthera subglobosa, breviter rostrata. 
At 6000 feet {Ilaviland). 

Another species of the Apodse section, without any dilatation of the stem into pseudo- 
bulbs, differing remarkably from all others in the very long slender scapes and the very 
short-sepaled flowers. The little flowers have the thin texture of the section, and are 
apparently yellowish. 

BuLBOPHYLLUM ALTISPEX, Eidl., n. sp. Caulis ascendens, ramosus, radicibus tenuibus, 
copiosis. Pseudobulbi subremoti, J poll, longi, cylindrici, vaginis striatis tecti. 
Polia elliptica, obtusa, carinata, J poll, longa, J poll, lata, carnosa. Eacemi brevissimi, 
paucifiori, bracteis tecti. Plores parvi, rosacei. Bracteae parvse, ovatae. Sepalum 
posticum oblongum, obtusum, J poll, longum, lateralia ovata, triangularia, subacuta, 
multo latiora. Petala linearia, ^ sepali sequantia. LabeUum linguiforme, marginibus 
ad basin elevatis tenuibus, apice carnosulum, sepalis brevius, brevissime stipitatum. 
Columnse stelidia securiformia, apicibiis productis acuminatis, antheram obtusam 
superantia. 

Epiphytic. At 8000 feet (Hamland, 1143). 

Plowers pinkish. Near B. capitatum, Lindl., but with blunter sepals and fewer flowers. 

Dendrochilum conopseum, Eidl., n. sp. Ehizoma longum, pseudobulbis cylindricis, 
remotis (1 poll, distantibus), poUicaribus. Polia elliplica, petiolata, 3 poll, longa, 
I poll, lata, petiolo J poll, longo. Eacemi graciles, rhachi tenui, 3-pollicares. Plores 
perparvi, dissiti. Bractese ovatse, acutee, pedicellos aequantes. Sepala linearia, obtusa. 
Petala subsimilia, angustiora. LabeUum breve, vix h sepaU aequans, panduratum, 



DE. O. STAFF OX THE FLORA OF MOUNT KINABALU. 237 

a basi ad medium inter carinas duas canaliculatum, ajuce ovatum. Colmniia brevis^ 
laciniis lateralibus e parte superiore longis, lincaribiis, inciirvis. 
Maripari Spur, alt. 5500 feet [Haviland). 

Eeia Scohtechinii, Stapf.— Syn. E. K'nigii, Hook. fil. EI. Brit. Ind. v. 7f)0 ; Ic PL 
t. 2066, non F. Muell. ; E. major, Kidl. Caulis vnlidus. Folia lineari-lanccolata, 
oblique acuta, acuminata, 15 poll, longa, f poll. lata. Scapi plures (ad 12), 
laniiginosi, validuli, bracteolis parvis, oralis, reflexis. Plores j):iiviili, di-H.-xi, 
lanuginosi. Ovaria ] poll, longa. Sepalum posticum oblongum, lateralia ovata, 
lanuginosa. Petala linearia, glabra. Labellum ungue brevi calloso, lobis lateralibus- 
brevibus, dentiformibus, medio ovato subacuto, callo tuberculo medio ; discus inter 
lobos laterales carnosus, elevatus, liaea mediana depressa. Capsula poUicaris^ 
lanuginosa. 

At 6000 feet (Hamland, 1250). 

Uatlier bigger than the Perak plant and with a slightly different lip, l)ut apparently 
specifically the same. 

Eria grandis, Ridl., n. sp. Caulis validus, f poll, crassus. Folia inferiora vaginanti 
superiora 18 poll, longa, J poll, lata, linearia, acuminata, iiiu3qualiter biloba, acuta, 
coriacea, multicostata. Scapi ex axillis superioribus orti 8 poll, longi, foliis brevi- 
ores, fere in tota parte exserta floriferi, rhaclii lanuginosi. BracteaB lanceolatte, 
acutee, J poll, longse, glabrae. Ovarium cum pedicello ^-1 poll, longuni, laiuigi- 
nosum. Elores pro sectione magni. Sepalum posticum lanceolatum, f \m\\. longuin, 
lateralia ovata, acuta, latiora, mentum rotundatum formantia, alba, pubescontia. 
Petala breviora -^q poll, longa, ovata, lanceolata, tenuiora, glabra. Labellum 
roseum, saccatum, quam petala brevius, marginibus anticis incrassatis, lobo medio 
brevissimo rostriformi. Columna rosea, majuscula, basi angusta, stelidiis erectis 
latis, apicibus dilatatis, clinandrio profundo reclinato. Antliera globosa, dorso fissa, 
emarginata. Stigma magnum, oblongum, callo parvo transverso ad basin columiue. 
Very common on the ground under jungle, from 10,000 to 12,000 feet [Havllandy 1157). 
This plant is very remarkable on account of the curious saccate lip, something in tlie 
shape of a coal-scuttle. The front edges at the entrance to the sac are thickened, and 
just within them is on each side another thickened low ridge; the thickened edges are 
continued into a short thick beak, which is all that represents the mid-lobe. 

Eria angustifolia, Ridl., n. sp. Caulis bipollicaris, foliis distichis equitantibus 
tectus. Folia subcoriacea, parte basali vaginante 2-3 poll, longa, lamina articulata 
6-15 poll, longa, \ poll, lata, lineari-acuminata, apice valde in^equilobo, lobo uno 
J-poUicari. Racemus brevior, vix 6 poll, longus, ex axilla folii superioris, multiflorus. 
Elores parvi, glabri, pedicellis \ poll, longis, minute et parce pubescentibus. 
Bractea? ovata?, reflexae, \ poll, longae. Sepalum posticum oblongum, obtusum, 
\ poll, longum, lateralia late ovata, 5-nervia, multo majora, ^ poll, lata ; mentum 
breve, latum. Petala ovata, subacuta, trinervia, dimidio sepali aequilonga. Label 
lum sepalo brevius, cymbiforme, lobis lateralibus triangidaribus erectis, lobo medio 
longiore triangulari subobtuso, si explanatum rliomboideum, carinis 4, 2 interi- 

SECOND SERIES. — BOTANY, VOL. IV. 2 I 



238 DE. 0. STAFF ON THE FLORA OF MOUNT KIXABALU. 

oribus in lobo medio conjunctis, exterioribus crassioribiis, sigmoideis, obscure 
pustulosis. Colunma brevis, marginibus versus basin incrassatis, papilla glandulosa 
terminata, clinandrii marginibus anticis rotundatis. 
At 6000 feet (Haviland). 

A curious species, the affinity of which is not clear. The stem and foliage recall those 
of MHa aeridostachya, but the glabrous flowers and short mentum make it very distinct. 
The lip is boat-shaped, with little erect lobes. On the walls of the boat are two thick 
sinuous ridges, thickest at the base, and on the terminal lobe, between them, in the middle 
line, are a pair of less elevated ridges which meet and form one on the mid-lobe. The 
foot of the column is somewhat abruptly bent, and the edges are thickened to form two 
ridges which end in a rounded glandular mamilla a little above the lip. It is probable 
that these ridges excrete nectar, witli which the boat-shaped portion of the lip is filled 
during life. 

Eeia ferox, Blume, Mus. Bot. Lugd. Bat. ii. 184. 

At 6000 feet (ITamlcmd). 

Spathoglottis aueea, Lindl., in Journ. Hort. Soc. v. (1850), 34 
Kinitaki River, alt. 4000 feet [Hamland, 1291). 

]NEPHELAPHYLLrM LATiLABRE, Hidl., n. sp. Rhizoma ascendens, vaginis papyraceis 
tectum. Polia ovato-cordata, petiolata, subacuta, 2 poll, longa, IJ poll, lata, petiolis 
J poll, longis. Scapus bipoUicaris, triflorus. Bractese magnse, papyracese, oblongse, 
\ poll, longse, pedicellis ccquilongse. Sepala lanceolata acuta, ^ poll, longa, ^ poll, 
lata. Petala linearia, subsequilonga. Labellum calcare brevi, recto, f poll, longo, 
ungue oblongo, limbo late reniformi, emarginato, f poll, lato, in disco versus apicera 
lineis tribus papillosis. Columna lata, basi attenuata. Anthera plana. 
At 6500 feet {Haviland, 1165). 
Lip pink-veined, with oblong orange centre. 

This species is allied to N. jmlchrmn, Blume, but differs in the broad reniform end of the 
lip, which has also three lines of small papillae in the middle, increasing in size towards 
the apex. 

CcELOGYXE PAPiLLOSA, Bidl., n. sp. Bliizoma lignosum, crassum; pseudobulbi elongati, 
cylindrici, 8-pollicares. Folia lanceolata, longe petiolata, acuta, 7-nervia, 8 poll, 
longa, 1 poll, lata, petiole 3-pollicari. Scapus e pseudobulbo baud evoluto, foliis 
partim evolutis, validulus, erectus, 15 poll, longus. Bacemus flexuosus, 4 poll, 
longus. Flores ad 6, mediocres, dissiti. Bracteae oblongae, ovatse, subacute, 1 poll. 
longse, J poll, latae, striatse (siccse brunneie), pedicello cum ovario paulo longiores. 
Sepala lanceolata, f poll, longa. Petala alba, sul)similia, angustiora. Labellum 
oblongum, sepalis subaequilongum ; lobi laterales parvi, obtusi, vix distincti, medius 
ovatus, disco dense papilloso, carinis sinuatis ad 9 inter lobos laterales. Columna 
basi angusta, superne alis magnis, triangularibus, subacutis, clinandrium superan- 
tibus. Anthera plana, ovata. 



■Dli. O. STAFF ON THE FLOKA OF MOUNT KINABALU. 239 

At 10,500 feet (Hamlanrl, 1098).* 

The stiff, tall scape, with a few flowers at the apex, the short, papillose lip, :iiul the lari;e 
bracts are peculiar marks of this species. 

Calanthe ovalifolia, Ridl., n. sp. Eolia late ovata, petiolata, acuta, lamina poll, 
longa, 5 poll, lata, glabra, petiolo 5 poll, longo. Scai)iis elatus 3-petlalis, pub(\sceiis ; 
racemus multiflorus, laxus. Flores parvi. Bracteae laiiceolata3, acuta?, J poll. longsR, 
pubescentes. Pedicelli fere 1 poll, longi, pubescentes. Sepala ovata, reflexa, ] ])oll. 
longa, 4 poll, lata, pubescentia. Petala obovata, spathulata, glabra. Labellum late 
ovatum, integrum, sepalis subtequilongum, ad basin laniellis duabus pubescenti bus- 
in structum. Calcar sepalis brevius, vix | j^oll. Ion gum. 
Penokok River, alt. 3000 feet {Ilcwilmid). 
One of the very few species with entire lij). 

Calanthe parviflora, Lindl., in Paxt., Fl. Gard. iii, (1852-1853), 37. t. 01. 

At 7000 feet {Haviland). 

" Corolla white, lip with a central orange spot." 

This plant fits very well with the short description in the * Folia Orchidacea,' wliich 
Avas based on a plant collected by Lobb in Java. 

Bromheadia rigida, Eidl., n. sp. Caulis validus, ultrapedalis. Folia subdisticha^ 
coriacea, linearia, o])tusa, apice emarginato, 5 poll, longa, f poll, lata, vaginis 
teretibus, ^ poll, longis, ore integro. Flores terrainales pauci in racemo brevissirao. 
Bractese magnae, late ovatse, coriaceoe, striatoe, fere ^ poll, longic. Pedicelli bracteis 
sequilongi. Sepala lanceolata, acuta, ^ poll, longa. Petala breviora, paulo latiora* 
Labellum trilobum, lobis lateralibus angustis, acutis, medio oblongo-obovato, margine 
crispo, disco elevato trinervio. Columna erecta, gracilis, alata, cUnandrii margine 
sinuato. Bostellum planum, late semiovatum. 
At 6000 feet [Bcwiland, 1251). 

" Sepals and petals exteriorly, column anteriorly with purple markings, lip with ill- 
defined yellow centre." A very distinct species, with stiff, narrow leaves. Most nearly 
allied to an undescribed species from Mt. Ophir. 

Appexdicula congesta, Eidl., n. sp. Caules 8 poll, longi, teretes. Folia tenuia, oblonga 
obtusa, multinervia, apice insequaliter bilobo, lobis obtusis, vaginis I poll, longis, 
superne dilatatis, papyraceis, ore integro. Bacemus terminalis, capituliformis, 
compactus, -| poll, longus, multiflorus. Bractea? flabellatge, fimbriatse, papyraceae, 
pallidiB, i poll, longge et latae (superiores minores). Flores ])arvi. Sepala ovato- 
lanceolata, acuta, lateralia mucronata, alba, exsiccando fascia mediana rubra. Petala 
angustiora et breviora, lanceolata, mucronulata. Labellum liberum, carnosum, 
ovatum, acutum, sepalis bre\ius, disco elevato, hippocrepiformi, ad basin rubro, callo 
parvo albo in apice labelli; calcar magnum, scrotiforme. Columna brevissima. 
Anthera tenuis, lanceolata. Pollinia 6, elongata, clavata, tenuia, translucentia, disco 

* From 8000 to 10,000 feet (Low).— Staff. 

2i2 



240 r>E. U. STAFF O^ THE FLORA OF MOU^T KINABALU. 

teniii, cuncato, magno, apice emarginato, snbtus ad apicem glandula rubra, bicruri, 
crassa instructo. Rostellum latum, tenue, oblongum, emarginatum. Stelidia acuta, 
erecta, rostellum suba^quantia, carnosa, rubra. Stigma latum. 
Penokok River, alt. 11,000 feet {Haviland, 1302). 

This is a very singular species, belonging to the series with strictly terminal inflores- 
cence. The raceme forms a compact cone Avith large, pale, fimbriate bracts. The flowers 
project but a short way, and appear to have been white with a median red bar on each 
sepal and petal, and a red horseshoe-shaped patch on the lip. The poUinia, too, are very 
curious ; they are 6 in number, very slender, and fixed in the middle of a thin- textured 
disc, very large in proportion to them ; at tlie end of this disc is a j^-sliaped, thick, red, 
gland-like body, which doubtless serves to attach it to the head of the fertilizing insect. 

Habenaria borneensis, RidL, n. sp. Caulis 2-pedalis, undique foliatus. Tolia ad 10, 
dissita, lanceolata, acuminata, cauli appressa, vaginanlia, 1-1| poll, longa, f poll. lata. 
Racemus laxus vel congestus, multiflorus, bracteis ovaria sequantibus, lanceolatis, 
acuminatis, J poll, longis, fere J poll, latis. Plores mediocres, virides. Sepalum 
posticum lanceolatum, \ poll, longum, erectum, lateralia lanceolata, obtusa, deflexa. 
Petala erecta, lanceolata, sepalo postico breviora et angustiora. Labellum lineare, 
carnosum, integrum, f poll, longum, ad basin lobis erectis, minutis, triangularibus. 
Calcar sequilongum, teres, obtusum, pendulum. Columna lobis stigmaticis, brevibus, 
rotundatis. Antherarum apices hand 2)roducti. Pollinia elongata, piriformia, pedi- 
cello tenui, disco ovato. 
At 10,000 feet [Havilaml). 
. One of the very few simple-lipped Asiatic Habenarias. The two little erect processes 
at the extreme base of the lip, however, doubtless represent the lateral lobes. 

SCITAMINEiE. 
Alpinia, sp. aff. A. ^nuticce, Roxb. 
At 5000 feet {Haviland, 1233). 
This is probably a new species, but the specimen is too young for description. 

H^EMODORACE/E. 
Aletris foliolosa, Stapf, n. sp. Herba 1^-2 ped. alta, glaberrima, rhizomate brevi. 
Eolia Humerosa, ad scapi hasin denslssime conferta, lineari-lanceolata, 1|— 3 poll, 
longa, basi 2-3J lin. lata, longe tenuiterque acutata, plana vel subplicata vel 
margmibus involutis, saepe leviter recurva, basi et ssepe superne distincte nervosa, 
9~ll-striata. Scapus 1-1 J ped. altus,yb///.s' 15-20 anguste linearihiis vel fil?formibus, 
4!-2 lin. longis, basi gibbosis, bracteiformibus obsitus, in racemum 3-4^ poll, longum 
abiens ; bracteie filiformi-lineares, 2|-1^ lin. longa? ; pedicelli ad medium vel i)aulo 
supra bracteola brevi obsiti, 1-2 lin. longi. Perianthium alljidum, tubuloso- 
campanidatum, ad medium ovario adnatum, 6-fidum ; lobi siibsequales, subspathu- 
lato-Jineares vel lineares, 1|-1| lin. longi. Eilamenta basi lobis adnata, iisque 



mi. O. STAFF ON THE FLORA OF MOUNT KJXABALU. 241 

dimidio breviora ; antherae j^urpurese. Ovarium somi-inferinii, ])artc libera fere 1 lin. 
alta; stylus f-1 lin. longus, suhtemus -, stigma distiiicte incrassatum. Cajisula 2-2^ 
lin. loDga. Semina lineari-oblonga, temiiter striata, vix h lin. longa. 
Maripari Spur, alt. 5500 feet : Temburungo, alt. 7700 feet (Jlaviland, 1125 r^ 1125 b). 
Distribution : Malaya, Sumatra, North Borneo. 

Well marked by the very numerous, densely crowded, and comparatively short leaves, 
and the numerous reduced leaves on the scape. It Avas collected by Bcccari also in 
Sumatra on Ht. Singalan (no. 167). 

Aletris rigida, Stai)f, n. sp. Herba 5-7 poll, alta, glaberrima, rhizomate brevi. i'olia 
numerosa, ad scapi basin densissime conferta, linearia vel lineari-lanceolata, 1-2 poll, 
longa, basi 1^-1 lin. lata, longe acutata, 2)lana vel subplicata vel marginibus 
involutis, rigide crecta, crassiuscula, basi excepta subenervia. Scapus 3-5 poll, altus, 
rigidiis,folUs hracteiformihus 2 lanceolato-Unearl-subnlatls 4-6 lin. longis obsitus, in 
raceraimi 2-1 poll, longum abiens ; bracteoe lanceolato-linearos, 4-2 lin. Jotigcp ; 
pedicelli brevissimi, basi bracteola brevi iis longiore obsiti. Perianthium albidum, 
tubuloso-campanulatum, ad medium ovario adnatum, 6-fidum; lobi linearcs vel 
lineari-oblongi, 1-1^ lin. longi. Eilamenta lobis basi adnata iisque dimidio breviora* 
Ovarium semi-inferum, parte liberti, vix ^ lin. alta, deinde magis protrusa; stylus 
brevis, crassiuscnlus, \ lin. longus; stigma paulo incrassatum. Semina lineari- 
oblonga, tenuiter striata, vix -| lin. longa. 
Erom 8000 to 11,000 feet [Low) ; in open damp places at 11,500 feet {Haviland, 1125 c). 
Closely allied to A.foliolosa, but distinct by the rigid habit, the few reduced leaves on 

the scape, and somewhat smaller flowers, with a shorter and stouter style. 

IRIDEiE. 

Patersonia Lowii, Stapf, n. sp. fPl. XX. b. 7-9.) Planta 8-12 poll. alta. Caulis 
brevissimus. Eolia circiter 12 cum scapo quoque, equitantia, linearia, longiora, 
7-11 poU. longa, 1^-lJ lin. lata, acutissima, planissima, rigida, in marginibus 
brevissime scarioso-fimbriata, glaberrima, tenuiter sed distincte multistriata. Scapus 
4-6 lin. longus, fere ad inflorescentiam folii intimi vagina amplexus. Spathae 2, 
subaequales, lineari-oblongae, acutse, IJ poll, longae, 2 lin. latae, striatae, late scarioso- 
marginatse. Bractese spathis similes, angustiores brevioresque, in carina dorsali plus 
minusve scarioso-fimbriatulse, interdum una alterave diminuta vacua addita. Peri- 
gonii tubus gracillimus, 11 lin. longus, vix exsertus ; lobi exteriores elliptico-obovati, 
5-7 lin. longi, circa 4-5 lin. lati, ut videtur, csBrulei vel purpurei, interiores nulli (?). 
Eilamenta in tubum integrum connata ; anthera) 1^ lin. longse. Stylus superne 
sensim pauloque incrassatus ; lobi stigmatici obovato-spathulati, minute crenulati. 
North-west face of Moimt Kinabalu, alt. 5000 feet {Loic) ; Maripari Spur, alt. 5500 feet 

{Saviland, 1259). 

I find it very difficult to point out the -affinity of this if compared with the 
Australian species. It may, perhaps, be considered as linked with P. glauca, U. Br., by 
wav of the species next to be describad. 



242 T)R. O. STAFF ON THE TLOEA OF MOUNT KINAEALIT. 

Patehsonia BORNEEysis, Stapf, n. sp. (PI. XX. A. 1-6.) Plaiita 1J-2J peel. alta. 
Caulis brevissiraus. Polia circiter 10-12 cum scapo quoqiie, eximie disticlia, 
equitantia, linearia, longiora, ad 2 j^ed. alta, ad 3 lin. lata, sequaliter acutata, 
planissinia, rigida, glaberrima, in marginibus levia et subincrassata, tenuiter sed 
distincte striata. Scaj^us foliis aequilongus vel paiilo brevior, in tertia parte 
superiore nudus. Spatha? 2 lineari-oblongae, acuta), striatse, scarioso-marginatse, 5-6- 
florae, inferior 16-18 lin. longa, 2^-3 lin. lata, superior 18-22 lin. louga, internodio 
distincto separata. Bracteae spathis consiniiles, paulo longiores, longe tenuiterque 
acuniinatae, scariosae. Perigonii tubus gracillimus, 1 poll, longus, breviter exsertus ; 
lobi exteriores, elliptico-oblongi, albi, purpureo-suffusi, 8 lin. longi, 4 lin. lati, 
interiores nulli. Pilamenta in tubum integrum connata ; antlierse 2 lin. longae. 
Stylus superne sensini leviterque incrassatus ; lobi stigmatici obovato-oblongi, 
margine papilloso-fimbriati. Capsula inter bracteas sessilis, f 230II. longa; valvae 
styli residuis caudataB. Semina ellipsoidea, 1 lin. longa, nigra, nitidissima. 
Temburungo, alt. 7700 feet (Eaviland, 1179). 

Allied to tlie preceding species, but mucli taller and with a somewhat different foliage. 
I have not seen the slightest trace of the inner row of segments in the perianth. Ttie 
structure of the spikes, which are few-flowered riphidia, is the same which is typical for a 
great number of Irideae, and amongst them for Sist/iHnchium, to which JPatet'sonia 
approaches indeed closely in more than one respect. 

LILIACE^. 

Smilax lj^vis, Wall. Cat. 5116 ; A. DC. Monogr. i. 56. 
At 7800 feet {miviland, 1115). 
Distribution : Malaya, in the Peninsula as far as Penang, North Borneo ; China. 

Smilax odoratissima, Blume, Enum. 19. 

Dahombang River, alt. 3000 feet {Mavilancl, 1293). 

Distribution : Malaya, from Java to the Philippines ; on the mainland to Burma and 
the Sikkim Himalaya. 

I include also >S'. aspericaulis, Wall., in this species, as the differential characters 
indicated by A. De Candolle do not prove trustworthy if tried with the material of 
>S'. asjjericaulis preserved at Kew. Prom the Pliilippines I have seen a form with 
emarginate leaves. It was collected in North Luzon ( Vidalj 1434). 

Smilax leucophylla, Blume, Enum. 18, a. et c. 

At 5500 feet (Havllcuid, 1226). 

Distribution : Throughout Malaya, from the Penmsula to the Philippines and the 
Moluccas ; Cochin-China. 

Dianella ensieolia, Ptedoute, Lil. t. 1. 

At 8000 feet ( Burhidge) ; Maripari Spur, alt. 5500 feet and 7500 feet (Raviland, 
1156, 1260). 



DE. O. STAFF OX THE FLORA OF MOUNT KINABALU. 243 

Distribution: Tropical Himalaya from Nopal eastward, Klinsia Hills, Muniporc, and 
Burma ; Ceylon ; Malaya to the Pacific ; Hascarenes. 

Burbidge's specimen represents the caulescent form with leaves 1 inch broad and 
glaucous, whilst Haviland's specimens possess leaves not more than 2-3J lin. broad and 
a much shortened stem. In no. 1156 (from an elevation of 7500 feet) the leaves are all 
basal except two very reduced ones on the scape ; in the other specimen there arc three 
distinct internodes visible, the middle one measuring 7 lines in length. Different as 
the habit of both forms may appear at the first glance, it is evident from the rich 
material which is preserved at Kew that both arc only different states of one species. 
It seems that the young branches, which always begin with a " rosette " of basal leaves, 
produce flowers sometimes in the first season, but more often in the second, when the 
stem has growm out and the leaves are separated by internodes from J to 1 inch. 

PALM^. 

Palms are repeatedly mentioned by the travellers, but there is not a single specimen 
in the Herbarium except an incomplete one of what I take to be a Fhmnga, collected 
by Lobb on the lower Tampassuk. Several species are grown near the Dusun 
houses, according to Mr. Barbidge, who mentions specially Cocos uncifera, hinw.y 
JPinangas^.,Arecas]).,Sind Oncosperma filamentostmiy Bliime. Another species oi Areca 
is stated by him to be wild, to measure about 10 feet in height, and to produce clusters 
of red fruits about 4 feet above the ground. " Rotang palms " seem to he common 
throughout the jungle of both mountain regions. Mr. Burbidgc figured two of them 
one apparently a true Calamus and the other a Korthalsia, with the foliage so charac- 
teristic of this genus. The latter was observed near Koung. 

AllOIDE^. 

Gamogyne Burbidgei, T^. E. Brown in Journ. Bot. xx. (1882) 195. 
In a ravine near Pakapaka {Burbidge). 
Distribution : North Borneo (Burbidge). 

EBIOCAULACE^. 

EniocATiLON HooKERiANrM, Stapf, n. sp. Caulis perbrevis. Folia densissime congesta, 
e basi latiore linearia, lJ-3 poll, longa, basi 2 lin., rarius ad 3 liu. lata, brevissime 
acutata vel obtusa, rigidula, erecta vel subfalcatim recurvata, vix pellucida (nisi in 
statu juvenili vel emarcida), tenuiter fencstrata, nervis circa 15. Scapus 5-7 poll, 
longus, sulcatus. Capitula 3 lin. dimetiens. Bractese irwolucrantes rotundatae 
vel late ovato-rotundatse, scariosic, glabrce, demum ex apice varie lacerae ; bracteie 
interiores late obovato-cuneatse, obtusissimae, caeruleo-cinereaB, dorso sub apice albo- 
pilosse. If lin. longa?. Receptaculum villosum. Elores utriusque sexus mixti. Elos 
masculus : Perianthium biseriatum 3-merum ; segmenta seriei exterioris a dorso 
compressse antice libera, caeterura ad f connata, obovata, albo-pilosa, 1 lin. longa ; 
segmenta seriei interioris exteriore paulo longioris stipitatse tubuloso-campanulatae 
albai intequalia, oblonga, sub apice glandula nigra notata, intermedium ^ lin. longum, 
lateralibus vix duplo majus. Stamina 6 ; filamenta alternatim longiora. Ovarii 



244 DE. 0. STAPP ON THE FLOKA OF MOUNT KINABALU. 

I'udinientum 3-lobum, nigrum. Plos f emineiis : Perianthium exterius segmentis 
distinctis spatliulato-cuneatis, superne navicularibiis, apice pilosis, 1^ lin. longis, 
interius stipite nudo distans segmentis albis spathulatis dense albo-villosis, glandula 
nigra notatnm. Ovarium sessile ; stjli exserti. Semen ellipsoideumvelsubglobosum, 
§ lin. longum. 
Maripari Spur, from 5000 to 5500 feet [Loic, Burbidge\ Saviland, 115S) ; Tembu- 
rungo, from 7500 to 7700 feet (Ramland, 1153 (Us), 1204). 
Distribution : North Borneo ; Tonkin (?). 

Closely allied to Eriocaulon snhcaulescens, Hook, fil., from Ceylon, but differing in the 
very short stem, the generally broader leaves, which are also more rigid and obtuse, in 
the darker heads, the less copious indumentum of the bracts and flowers, and the less 
unequal Igbes of the inner male perianth, which does not very much exceed the outer, by 
which it is enveloped almost entirely. This is also compressed from the back and slit to the 
base in front, but otherwise connate to J at least. The specimens collected on Kinabalu 
are all very uniform, but those of Balansa from I'lle Yerte, east of the bay of Fi-Tsi-Long 
in Tonkin (246), have somevvliat larger heads, a more copious indumentum, and a longer 
middle lobe in the inner male perianth, thus approaching E. stibcatdescens, and I am 
not quite sure whether it can really be referred to E. Sookeriannm. On the other hand, a 
specimen collected by Beccari in Sarawak (2420) agrees perfectly with the Kinabalu plant. 

CYPEEACE^. 
SciRPTJS iNTJNDATUs, Sprcngel, Syst, i. 207. 

At 11,000 feet (HamlaiuU 1397). 

Distribution : North Borneo ; Australia from Queensland to Tasmania ; Norfolk Island ; 
New Zealand. 

SciRPUS Clakkei, Stapf, n. sp. Csespitosus, J-2 ped. altus, gracilis. Ilhizoma breve, 
dense fibrosum, culmos numerosos, dense fasciculatos, edens. Culmi ^-2 ped. alti, 
teretes, sulcato-striati, tenues, crassiores, vix, \ lin. dimetientes, ad 2-4 poll, vaginati. 
• EoKa inf eriora ad vaginas tenuiter striatas, fuscescentes, arete culmo adpressas, oblique 
truncatas, mucrone -|-1^ lin. longo superatas redacta, simimum interdum longe 
vaginans lamina suhulatd ^-1 poll, longd, plus minusve plicata, marginibus serrato- 
denticulatis. Spiculae solitaries, oblongae, lineares, 3-4 lin. longte, pauci- vel ad 
8-florae ; bracteae paucse (interdum 5), infimae vacuaj, caeteraj fertiles. Bractea infirna 
a cseteris internodio distincto remota, late rotundata, in mucronem brevem viridem in 
margine scabrum excurrens, sequentes inter se sequales, ovato-oblonga?, 1^-lf lin. 
longse, obtiiscB vel subobtusse, muticae, pallide castaneae, carinatse, carina viridi. 
Perianthii setae tenuissimoe, tiexnoste, 1-1^ lin. longse, breviter parceque papillosa?, 
(an semper ?). Stamina 3 ; lilamenta linearia, 1^ lin. longa ; antherue vix f lin. 
iongse, in fioribus superioribus plus minusve redactse. Stylus in fioribus inferioribus 
ad 5^- J, in superioribus ultra medium in ramos 3 stigmatosos, filiformes lissus. 
Caryopsis cinereo-fusca, obovato-oblonga vel oblonga, ad 1 lin. longa, apice subito 
a,piculata, dorso convexa, leviter carinata, facie subplana. 
At 11,000 leet {Raciland, 1398). 



DR. 0. STAFF OX THE I'LOEA OF MOUNT KIXABALU. 245 

Allied to Seirpus subcapitatiis, Tiiw., but distinct by the very slender habit, the more 
advanced development of the lamina of the leaves, and spikelets always solitary, with 
rather obtuse bracts. It approaches very closely certain Chinese specimens which 
probably belong to S. suhcapitatus, the fruit of wliich, however, is still unknown,. 
particularly one from the lower Yang-tse-Kiang (Maries), and also a si)ecimen collected 
by Beccari on Mt. Singalan in Sumatra (no. 276). However, the latter is much stouter, 
and the bracts of the spikelets are longer and more acute. Mr. C. B. Clarke considers 
it to be also a form of S. suhcapitatus. 

ScHCENLS APOGON, EcBmcr et Schultes, Syst. ii. 77. — Syn. Chcctospora imberbis, 11. Brown,^ 
Prodr. 233 ; S. Brownii, Hook. fil. Handb. N. Zeal. PL 298 ; Bentham and F. Mueller, 
EL Austral, vii. 373. 
In crevices of rocks from 11,000 to 13,000 feet {Ilaviland, 1395, 1396; Loic at 
9000 feet). 

Distribution : From Queensland to Tasmania ; North Borneo ; Loo-Choo Islands. 

ScHCENUS MELANOSTACHYUS, U. Brown, Piodr. 231. 
Maripari Spur, at 5500 feet {Haviland, 1407). 
Distribution : Queensland to New South Wales ; West Australia (?) ; North Borneo. 

Cladium borneense, C. B. Clarke, n. sp. C. " culmis robust ioribus, terctibus; foliis suh- 
nullis, paniculd oblongd, densd, spiculis composite fasciculatis ; ptistillo juuiore ellip- 
soideo-tt^igono, rostro ovoideo-conico, glabro, stylo S-Jido. Rhizoma repens, t mm. in 
diam., ligneum. Culmi metrales, 3 mm. in diam., nisi basi nudi. Vagina summa 1 dm. 
longa, in altero latere 1—1 cm. lanceolato-producta. Panicula 6 cm. longa, 2 cm. 
lata ; rami approximati, rigidi, anfractuosim flexuosi, suberecti, minute scabridi ; 
bractea ima vix vaginata, cum |-|- parte panicula? suba3quilonga. Spicula) 6 mm. 
longa?, fusee brunnese, sub 5-gluma?. Glumae undique imbricatse, ovato-oblongse, 
l-nervise, minute scabro-puberuloe ; 3 imte vacuae, apice obtusa?, subaristellatae, 
quarta perfecta, nucifera, cum imis a^quilonga, vix aristellata, summa angustior, 
obtusa, vacua. Setae 3, minuta?, ovatae. Stamina 3; anthera? lauceolato-cristatae. 
Stylus Unearis, glaber ; rami 3 lineares." 

At 5000 feet (Lotc). 

Distribution : North Borneo [plains near Tamparluli, in moist places (Burbidge)]. 

" Species C. vagiuali proxima, sed inflorescentia fere Cladii Vauthericc ; etiam a 
C. teretifolio, U. Br., parum distat." 

Cladium samoense, C. B. Clarke, n.sp.— Syn. Vincentia latifolia, Nad. PI. Tahiti, 34(?), 
non K-unth. C. " panicidd laxiusculd ; spiculis atrorubris ; glumis minute scabro-pube- 
rulis ; stylo S-Jido ; nuce triquetral in stipite lineari-obconico sustentd, rostro cum nuce 
cequilongo, scabro-pilosulo. Characteres sectionis Vincentice. Culmus 6 dm. longus. 
Folia 12 mm. lata. Panicula 2 dm. longa, fere 1 dm. lata ; ramuli scabro-pilosuli. 
Spicule 4-5 mm. longae. Filamenta cum stylis exserta. Nux vix alata ; stipes 
nuce propria multo brevior, obpyramidalis." 

SECOND SERIES. — BOTANY, VOL. IV. 2 K 



246 DE, O. STAFF ON THE FLOEA OF MOUNT KINABALU. 

Maripari Spur, at 5500 feet {Savilcmd, 1406) ; at 11,500 feet {Saviland, 1405). 

Distribution: Polynesian Islands (Viti, Samoa, Tahiti). 

" In exemplo Bidwellii, planta major, panicnla magis evoluta, nux non matnrata." 

[ISToTE. — Mr. Clarlie distinguishes a variety ^^unlseta; seta unica, cum f parte nucis 
subaequilonga, filiformi, levi, rubra." The Borneo specimens possess a similar, but pale 
seta ; their nuts, however, are more elongate in all parts. — O. Stapp.] 

Oahex rara, Boott, Trans. Linn. Soc. xx. 139. 

At 11,000 feet {mwilcmd, 1393). 

Distribution : East Himalaya and Khasia HiUs ; Ceylon. 

A very closely allied species, C. capillacea, Boott, is found in the Sikkim and Bhotan 
Himalaya, in Manchuria, Japan, and in New South Wales. 

Caeex FUsiroEMis, Nees, ex WigJit, Contr. 128. 
At 11,000 feet {Baviland, 14^04^). 
Distribution : Temperate Himalaya, from Sikkim to Kumaon. 

€akex hypsophita, Miquel, El. Ind. Bat. iii. 354. 
At 10,000 feet {Hcwiland, 1403). 
Distribution : Malaya, from Sumatra to Java. 

Cahex pilicina, Nees, ex Wightj Contr. 123. 
At 10,500 feet {Saviland, 1402). 
Distribution : Himalaya and Khasia Hills ; S.W. Deccan and Ceylon ; Malaya. 



GRAMINE^. 

IsACHNE Ktjnthiana, Necs, ex "Wight & Arnott, Cat. n. 1659. — Syn. JPanicum Kwithianum, 
Wight & Arnott ex Steud., Syn. Gram. 96. 
At 8000 feet {Haviland, 1408). 
Distribution : Ceylon and S.W. Deccan ; Malaya. 

OpLisMExrs coMPOsiTrs, Beauv., Agrost. 54. 

Dahombang River, at 3000 feet {Haviland^ 1409). 

Distribution : As a weed, widely spread throughout the Tropics. 

MiscANTHUS SINENSIS, Audcrsson in Ofvers. Vet. Akad. Forhand. Stockholm, 1855, 
p. 166. 

At 5000 feet (Low). 

Distribution : China to Tonkin and to Corea ; Loo-Choo Islands and Benin Isl. to 
Japan. 

Aghostis canina, linn. Spec. Plant ed. I. 62. 

var. BORNEENsis, Stapf, n. var. Culmi longe ascendentes, ad vel ultra paniculam 

laxe vaginati. Panicula contracta, 2 poll, longa, atropurpurea, folio summo supe- 



J)E. O. STAFF OK THE FLOE A OF MOUNT KIN AB ALU. 247 

rata ; rami ramulique fere laeves. GliimaD vacua- suhrrquales, 1^-1 1 lin. lon"-a3^ 
florens 1 lin. longa, rigidior, ex medio vel iiimlo supra arista ta ; arista J] lin. longa. 

At 13,000 feet {Savilmid, 1399). 

Distribution (of species): Europe, with tlie exception of the Mediterranean province^ 
Cavicasus and N.E. Asia Minor; Southern Siberia to Kaslimir and Sikkim : North 
America from Greenland to the North-eastern States; Aleutian Islands to Sitka; 
Falkland Islands and Southern Patagjonia ; Ncav Zealand and Campbell Island. 

I have very carefully compared this grass with the numerous varieties of Agrostis 
canina, but I have not been able to identify it exactly with any one of them. It 
certainly approaches nearest the variety which was originally described as A.falklandica 
by Sir J. Hooker, and a form collected by Colenso in New Zealand (no. 4142), so far as 
the general habit is concerned. But it has decidedly larger spikelets, with a more rigid 
flowering glume than the former, whilst the awnless spikelets of the latter are 2 lin, 
long. The form found in the Sikkim Himalaya deviates more in the habit and in the 
small awnless flowering glume. 

Deteuxia epileuca, Stapf, n. sp. (PL XX. c. 10-lG.) Gramcn perenne, cyespitosum, 
\\-'2i poll, altum. Culmi floriferi atque innovationes intravaginales basi arete vagi- 
nati, illi (demum quidem) a medio vel ultra nudi, hae minute pubescentes. Foliorum 
vagina glabra, prominenter striata, ligula obsoleta, lamina linearis plana vel in statu 
juvenili vel Immefacta plicata, |-li poll, longa, -^-| lin. lata, obtusiuscula, in mar- 
ginibus scabridula, siqyni albo-glauca, suhtus pallkle virldis, vix glaucescens, propter 
nervum medium carinatim prominenteni nervis utrinque 3-4 plus minusve prominulis 
percursa. Panicula parva, oblongo-ovata, 3-6 lin. longa, vix 3 lin. lata ^el anguste 
contracta ; ramuli breves, pauciflori, stricti, dense hispiduli ; pedicelli interdum 
brevissimi. Spiculse l-florae; rhachilla glabra ultra Jlorem in setam glahram strictam 
\ lin. longam produota. Glumae vacuas 2, ovatas, a latere comprcssa?, carinatje, 
acutae, in carina hispidulae vel glabratae, nervis lateralibus abbreviatis, sa3pe indis- 
tinctis, utrinque 1, rarius 2, superior paulo longior, f-f lin. longa ; florens oblique 
oblonga, 1 lin. vel paulo ultra longa, obtuse subcarinata/, laevissima, subcartilaginea, 
trinervis, nervis lateralibus sub margine paulo inflexo fere ad apicem ductis, stepe 
nervulis tenuissimis, utrinque 1, interjectis. Palea aequilonga obtuse subcarinata,^ 
Isevissima, dorso profunde sulcata. Lodiculse 2 ad medium connata?, sequales vel 
subaequales, ovario paulo breviores. Stamina 3 ; antherae vix \ lin. longaj. Ovarium 
ovoideum ; styli breves, distincti ; stigmata plumosa. Caryopsis oblonga, subobliqua, 
gluma paleaque inclusa, libera, | lin. longa ; hilum basale, obliquum. 
At 10,5000 feet {Saviland, 1401). 

This is a very marked species, the affinity of which lies rather with some Australian 
species than with any others, though it is far from bemg closely connected. The general 
habit is like that of dwarf specimens of J>. minot^, Benth., but the glumes are smaller and 
much less acute, the flowering glume is firmer and awnless, and the rhachis distinctly 
produced. The spikelets come, perhaps, nearer to those of D. G-unniana, Benth., but this- 
is a very delicate and probably annual grass with setaceous leaves. 

2k2 



248 pn. O. STAFF ON" THE FLOEA OF MOUNT KINABAXU. 

Deschampsia iLExuosA, Trinius, in Bull. Sc. Acad. Petersb. i. (1836), 66. 

' var. LiGULATA, Stapf, n.Yar. Eolia risjidiora, crassiora; ligiilso foliorum intermedi- 

orum 2-4 lin. longse, albo-scarlosa?, tenues, fere ad basin bitidse. 

At 13,000 feet {Havilcmcl, 1400). 

Distribution (of species) : Europe from Scandinavia to the Mediterranean Sea and to 
the Caucasus ; North America from Greenland to Canada and the Northern United 
States ; Falkland Islands and Magellan Strait. 

A specimen from Bale Frangaise, in the Falkland Islands, comes nearer to this variety 
than any other specimen I have seen. The leaves measure over J lin. in diameter, but 
their anatomical structure is exactly of the same type as that of the common form with 
setaceous leaves. The difference is due to the enlargement of the cells, whilst the greater 
stiffness is caused by the stronger development of the subepidermal layer of sclerenchyma. 
The awn is straight in the Kinabalu specimens, but they are still young ; and the awns 
of typical D. fleccuosa may also be found to be so in this state. 

ScnizosTACHTUM Blumii, Nees, Agr. Bras. 535 (?). 

Taw-aran, at 1600 feet {Bavilcmd, 1387). 

Distribution : ISIalaya, from Java to Borneo. 

The specimen is too imperfect to make the determination perfectly safe. It seems, 
how^ever, to be certainly identical with specimens collected by Motley near Banjermassin. 
One of these was named S. hraehycladum by Munro in MSS. ; but he added " very near 
S. Blinnei" and then " S. Blumei ?," and I cannot distinguish it from the present species. 
But these specimens, again, are not in a sufRciently perfect condition for settling the 
question. 

DiNOCHLOA (?), sp. 

Penokok River, 3000 feet (Saviland, 1390). 

The flowers are too young for description, but the plant has perfectly the appearance 
of D. Tjankon^eh, Buese, with the exception that the spikelets are larger and each 
cluster is supported by a bract of almost equal length. 

CONIFERS. 
Dacrydium elatum, Wallich, Cat. 6045 ; Endlicher, Conif. 226. 
At 6,600 feet {Haviland, 1183). 
Distribution : Malaya, from the Peninsula and Sumatra to Borneo ; Cambodja. 

Dacrtdium, n. sp., aff. D. elato, Wallich. 

Small tree with pendulous branchlets, when young ; from 8000 to 11,000 feet {Low). 

A very characteristic species with long pendulous branchlets, with the leaves about 
\ in. in diameter. The densely crowded leaves are erect, but slightly curved outwards 
and j m. long and \ lin. broad. They are triangular in transverse section, with a 
distinct keel on the back and a faintly raised middle nerve above, and they terminate 
with a hard and sharp mucro. The specimens have neither flowers nor fruits. 



DE. O. STAFF O^^ THE FLOEA OF MOUNT KINABALU. 2iQ 

Phtllocladus hypsophylla, Hooker fil. Ic. PI 889. 

A tree 10-30 feet high at lower elevations, from 8000 to 10,000 feet (Low); a shrub 
from 10,000 to 12,000 feet {Havilcmd, 1092). 

Distribution : North Borneo ; New Guinea. This species was found also on Loboiii^ 
Peak, N.E. Borneo, at 4000 feet, by Lobh, and in Sarawak by Beccari (no. 2391, 3220). 

PoDOCARPUS CUPRESSINA, R. Brown, ex Mirbel in Mem. Mus. Par. xiii. 75. 

At 11,000 feet {Eca-lkuid, 1094). 

Distribution : Malaya, from the Peninsula and Sumatra to the Philippines ; Hainan ; 
North Burma; New Caledonia and Viti Isl. (?). 

The locality where Griffith collected the plant is near the serpentine mines south of 
the Hoo-Koom Valley, between 1600 and 2G00 feet. The New Caledonian specimens are 
not quite complete and may perhaps belong to another species. 

PoDOCAEPUS NERiiFOLiA, D. Don, in Lamb. Pin. edit. I. p. 122, et 3, p. 74 ; Hooker fil. 
PI. Br. Ind. v. 649. 

var. BREVIFOLIA, Stapf, n. var. Polia densissima, ^-1| ])oll. longa, 2-3 lin. lata. 

Amenta mascula solitaria. Pructus ovoideus, 5 lin. longus, receptaculum bracteis 
anguste ovatis integris sufFultum sequans. 
Prom 11,000 to 12,000 feet {Low, $ ; Raviland, 1093, 6 ). 

Distribution (of species) : Malaya, from the Peninsula and Sumatra to the Philippines; 
Tropical Himalaya and Khasia Hills ; Indo-China ; Central China and Japan, 

I have not been able exactly to match the present form, which comes nearest to the 
one described as Todocavpus polystachya by B. Brown, and merged by Sir Joseph Hooker 
in P. neriifolia. It is probably an alpine state of P. polystachya^ with a stunted habit 
owing to an elevated and exposed locality. 

PILICES. (By J. G. Baker.) 

Gleichenia circikata, Swartz, Syn. Pil, 165; Hooker & Baker, Syn. Fil. ed. II. 11. 
At 5000 to 6000 feet {Burbidge) ; at 10,500 feet (Baviland, 1483). 
Distribution : Prom Australia and New Zealand to Malacca and the Philippines. 

Ctathea Havilandii, Baker, n. sp. (37 *). Stipites validi, paleis lanceolatis copiosis 
brunneis crispatis ad apicem proediti. Lamina oblongo-lanceolata, tripinnatifida, 
subcoriacea, glabra, bipedalis, 9-12 poll, lata, infra in pinnularum costis dense 
paleacea ; rhachis valida, paleacea ; pinnae sessiles, oblongo-lanceolata', inferiores 
numeros£e, 5-6 poll, longae, 2 poll, latse, infimae baud reductoc; pinnuhe plurimae 
confert^e, lanceolatae, sessiles, inferiores profunde pinnatifidte, segmentis tertiariis 
lineari-oblongis, erect o-patentibus ; venulae simplices4-5-jugaB, erecto-patentes. Sori 
in pinnularum costis uniseriati. Indusium firmum, cupulatum, persistens, glabrum, 
ore truncate. 
At 10,500 feet {SavHand, 1485). 
A very distinct species, most resembling in texture and final segments the America a 

Alsophila pruniata. 

* The numbers attached to the new Ferns indicate their position according to the sequence of species followed iu 
Hooker & Baker, ' Synopsis Filicum.' 



250 DE. O. STAPE ON THE FLOEA OF MOUNT KINABALU. 

Cyathea poltpqda, Baker, n. sp. (38). Lamina ampla, subtripinnata, utrinque viridis, 
glabra, infra paleis copiosis albidis ovatis convexis, praesertini in segmentorimi costis, 
prsedita ; rliacbis nucla, inermis ; pinna? oblongo-lanceolatse, sesquipedales, 4-5 poll, 
latse ; pinnulse lanceolatse, pinnarum apice excepto distincte petiolatse, inferiores 
2-2^ poll, longse, 6 lin. latse, psene ad costam pinnatse ; segmenta tertiaria oblonga, 
integra, obtusa, lJ-2 lin. lata ; venae T-S-jugse, venulis profunde f arcatis. Sori inter 
segmentorum costam ct marginem mediales. Indusium capulatum, parvum, 
glabrum. 

At 7000 feet {Savilcind, 1479). 

Near Cyathea integra, J. Sni., a native of Amboina and tbe Philippine Islands. 

Alsophila latebeosa, Hooker, Spec. Pil. i. 37 ; Hooker & Baker, Syn. Eil. 43. 
Thirty feet high, common at 2000 feet (Zoic). 
Distribution : Common almost throughout India proper and Malaya to Formosa. 

HYMENOPiiYLLrM DiLATATUM, Swartz, Syn. Eil. 147 ; Hooker & Baker, Syn. Pil. 62. 
From 5000 to 7000 feet (Burbidge). 
Distribution : From New Zealand and Polynesia to Penang, Tonkin, and the Philippines. 

Hymenophyllum Neesii, Hooker, Spec. Fil. i. 99 ; Hooker & Baker, Syn. Fil. 76. 
Without precise locality [Low, Burbidge). 

Trichomanes digitatum, Swartz, Syn. Fil. 370 ; Hooker & Baker, Syn. Fil. 76. 
Without precise locality {Burbidge). 
Distribution : Malaya ; Polynesia ; Mascarenes. 

Teichomanes PALLiDrM, Blumc, Enum. PL Jav. 225 ; Hooker & Baker, Syn. Eil. 80. 
Without precise locality {Burhidge). 
Distribution : Ceylon ; Malaya ; Queensland ; Samoa. 

TeichomajsES Eilicijla, Bory, in Duperrey's Voy. Bot. 283 ; Hooker & Baker, Syn. 
EH. 81. 

Without precise locality {Burbidge). 

Distribution: Widely spread throughout the Tropics of the Old World, and in 
Polynesia. 

Trichomanes microchiltjm, Baker, n. sp. (49). Bhizoma filiforme, late repens. Stipites 
graciles, nudi, erecti, lJ-2 poll, longi, sursum anguste alati. Lamina oblongo- 
lanceolata, glabra, decomposita, 3-4 poll, longa ; rhachis ad basin anguste alata ; 
pinnae sessiles, inferiores lanceolatse, polli cares, basi postice cuneato-truncatye, 
infimse hand reductse ; pinnulse infimae deltoidese, segmentis ultimis ereclo- 
patentibus anguste linearibus uninerviis integris. Sori pauci, terminalcs, sessiles. 
Indusium oblongum, f lin. longum, ore labiis' 2 minutis ovatis integris erectis 
membranaceis instructo. 

At 7000 feet {Kmlland, 1478). 

Habit and cutting of Hyraenophyllmn polyanthos, Sw., and H. demismm, Sw. 



DK. 0. STAFF ON THE FLORA OF MOUXT KINABALU. 251 

Trichomanes rigidum, Swavtz, Syn. Fil lil; Hooker & Baker, Syn. Fil. 86. 
Prom 5000 to 7000 feet {Burbidge). 
Distribution : Throughout the Tropics of both hemispheres. 

Teichomanes apiifolium, Presl, Hymenoph. 16 et 44; Hooker & Baker, Syn. iil. 86. 
At 6000 feet, on Casuarina trees {Burbidge). 
Distribution : Malaya ; Philippines ; Polynesia ; Norfolk Island. 

Trichomanes maximum, Blume, Enum. Fil. Jav. 228 ; Hooker & Baker, Syn. Fil. 86. 
From 5000 to 6000 feet [Low, Burbidge). 
Distribution : Malaya to North Australia and Polynesia. 

Trichomanes Pluma, Hooker, Ic. PI. x. 997. 
From 5000 to 6000 feet {Burbidge). 
Distribution : North Borneo ; Perak. 

Trichomanes trichophyllum, Moore, Gard. Chron. 1862,45; Hooker & Baker, Syn. 
Fil. mQ. 
Without precise locality {Low) ; from 5000 to 6000 feet {Burbidge). 
Distribution : Malaya ; New Guinea ; New Caledonia. 

Davallia alpina, Blume, Enum. PL Jay. 231 ; Hooker & Baker, Syn. Fil. 89. 
At 10,500 feet {Eamland, 1491). 
Distribution : Malaya ; Polynesia. 

Davallia contigua, Swartz, Syn. Fil. 130 ; Hooker & Baker, Syn. Fil. 94. 
At 10,500 feet {Hamland, 1484). 
Distribution: Ceylon; Malaya; Polynesia. 

Davallia Veitchii, Baker, Journ. Bot. viii. (1879) 39. 
From 6000 to 7000 feet {Burbidge). 
An endemic species. 

Davallia ciliata. Hooker, Spec. Fil. i. 184, t. 60 A. ; Hooker & Baker, Syn. Fil. 98. 
At Koung {Burbidge). 
Distribution: Philippines. 

Davallia te^^uipolia, Swartz, Syn. Fil. 130 ; Hooker & Baker, Syn. Fil. 102. 
At Koung {Burbidge), 
Distribution : Common in Tropical Asia and in Polynesia ; Japan ; Mascarenes. 

LiNDSATA Jamesonioides, Baker, Journ. Bot. viii. (1879) 39. 
At 5500 feet {Haviland, 1495) ; at 9000 feet {Burbidge), 
An endemic species. 

LiNDSATA flabellulata, Dryaudcr, Trans. Linn. Soc. iii. 41, t. 8. f. 2; Hooker & 
Baker, Syn. Fil. 107. 



252 DE. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

On tlie Dahombang River (Burhidge). 

Distribution : Malaya to South China, North India and Ceylon, and to Australia. 

Pteeis incisa, Thunberg, Prodr. PL Cap. 171 ; Hooker & Baker, Syn. Pil. 172. 
At 7000 feet {Savilmd, 1492). 
Distribution : Through the Tropics of both hemisjjheres. 

AsPLENiUM SQUAMULATUM, Blumc, Enum. PL Jav. 174 ; Hooker & Baker, Syn. Pil. 192. 
At Koung [Burhidge). 
Distribution : Malaya and Philippines. 

AsPLENirH BORNEENSE, Hooker, Spec. Pil. iii. 136, 1. 186 ; Hooker & Baker, Syn. Pil. 203. 
On the Dahombang Eiver, at 2000 feet {Loid) ; at 3200 feet {Haviland, 1477). 
Distribution : North Borneo ; Perak. 

AsPLENiuM LASERPiTiiEOLiUM, Lamarck, Eneycl. Bot. ii. 310; Hooker & Baker, Syn. 

Pil. 215. 
' Prom 2000 to 3200 feet (Haviland). 

Distribution : Prom Polynesia and North Australia to Chusan and the Bonin Islands ; 
Assam. 

AspLENiUM DiCHOTOMEM, Hooker, Spec. Pil. iii. 210 ; Hooker & Baker, Syn. Pil. 221. 
Without precise locality {Low) ; in shady jungle near Kiau {Burhidge). 
Distribution : North Borneo. 

AsPLENiEM BISERIALE, Baker, n. sp. (209) ; (§ Biplazium). Caudex breviter repens. 
Stipites erecti, li-2 poll, longi, ad apicem paleis lanceolatis squarrosis brunneis 
Testiti. Lamina obloDgo-lanceolata, simpliciter pinnata, glabra, utrinque viridis, 
4-6 poll, longa, lJ-2 poll, lata, rhachi paleis parvis vestita; pinnae multijugse, 
sessiles, lanceolatae, superiores integrte, iuferiores 1 poll, longse, 2 lin. latae, prof unde 
crenatse, basi inaequales, postice reductae ; venae obscurse, erecto-patentes. Sori 
costulares, in utroque latere costae uniseriati, lineari-oblongi, ssepe diplazioidei. 
Indusium latum, membranaceum, glabrum, persistens. 

At 3000 feet {Scmland, 1475). 

The specimen resembles a much reduced form of the cosmopolitan A. sylvaticum, Presl. 

AsPLENiUM LATiEOLiUM, D. Don, Prodr. Pi. Nepal. 8 ; Hooker & Baker, Syn. Pil. 239. 

In shady jungle near Kiau {Burhidge). 

Distribution : Ceylon and Nilgherries ; Sumatra, Borneo, and Philippines ; South 
China. 

DiDYMOCHLJENA LUNULATA, Dcsv. Prodr. 282 ; Hooker & Baker, Syn. Pil. 218. 
At 3500 feet {Burhidge). 
Distribution : Throughout the Tropics of both hemispheres. 

Nephrodium GYMNOPODLii, Baker, n. sp. (17) ; (§ Lastna). Stipites graciles, nudi, nitidi, 
fusci, 4-5 poll, longi. Lamina oblonga lanceolata, bipinnata, subcoriacea, glabra. 



DE. O. STAFF OX THE FLOEA OF MOOT XJXAJULU. 25^ 

semipedalis, rhachi utrinque jiubesceiite, hautl paleaceri ; pinnic libera?, circitcr 
20-jugae, iiifimic maximtc, deflexa), sessiles, lanceolatiu, 15-18 lin. long;o, 4 lin. latie, 
ad costam pinnatse, segmentis lineari-oblongis, integris, ^ lin. latis ; vena} remot:i\ 
ereeto-patentes, simplices. Sori ad segmentorum basin soli tarii, mediales. Indusiuni 
parvum, reniforme. 

At 10,500 feet {miviland, 1486). 

Belongs to the group of ]V. patens, Desv., near N. chrysolohum, Fee. 

Poi^YPODiLM Havilandii, Baker, n. sp. (107*); (§ Grammltis). Caules ca^spitosi, 
graciles, subpoUicares, pilis snbtilibus, mollibiis, patentibus tenuiter vestiti. Ljimina 
simplex, lanceolata, subcoriacea, utrinque pubescens, 2-3 poll, longa, medio 3 lin. 
lata, ad basin et apicem sensim attenuata, margin o obscure crenato; vena? obscurie 
furcatae. Sori oblongi, obliqui, uniseriati, laxe dis])ositi, ad costam eontigui. 
At 10,500 feet {JEaviland, 1488). 

PoLYPODiTJM cucuLLATTJM, Nees et Blume, in Nov. Act. Cses. Leop.-Car. Nat. Cur. ii. 121 , 
t. 126 ; Hooker & Baker, Syn. Fil. 324. 
Without precise locality (Loio) ; at 10,500 feet {Rariland, 1482). 
Distribution : Ceylon ; Malaya, Philippines and New Guinea ; Polynesia. 

PoLYPODiUM BLECHXOiDES, Hookcr, Sp. Pil. iv. 180; Hooker & Baker, Syn. Fil. 331. 
At 10,500 feet {Raviland, 1487). 
Distribution: Polynesia; Queensland. 

PoLTPODiUM CLAViFER, Hooker, Sp. Fil. iv. 176 ; Hooker & Baker, Syn. Fil. 334. 
Without precise locality {Low). 
Distribution : Borneo ; New Guinea. 

PoLTPODiUM TAXODioiDES, Baker, Jouru. Bot. viii. (1879), p. 42. 
On wet rocks on torrents {Burhidge). 
An endemic species. 

PoLTPODiTJM STENOPHYLLUM, Blumc, Enum. PL Jav. 124 ; Hooker & Baker, Syn. Fil. 
354. 
At 10,500 feet {Raviland, 1489). 
Distribution : Malaya and Philippines. 

PoLYPODiUM SOEIDENS, Hooker, Spec. Fil. v. 61, t. 283, b; Hooker & Baker, Syn, 
Fil. 355. 
At 8000 feet [JSaviland, 1493). 
Distribution: Borneo. 

PoLYPODiUM LACiNiATTJM, Blume, Enum. PI. Jav. 131 ; Hooker & Baker, Syn. Fil. 365. 
At 10,500 feet {Havlland, 1490). 
Distribution : Malaya. 

SECOND SERIES. — BOTANY, YOL. IV. 2l 



254 DE. 0. STAPr ON THE FLOEA OF MOUNT KINABALU. 

PoLTPODiUM COSTULATUM, Baker, Journ. Bot. 1880, 2L5. — Syn. Acrostichum costulatum, 
Cesati, Fil. Becc. Polyn. 8. 
On Daliombang Kiver (Haviland, 1474). 
Distribution : Malaya ; New Guinea. 

Gymnogramme avenia, Baker, in Hook. & Baker, Syn. Eil. 388. 
On rocks near Koinig [Biirbidge). 
Distribution : Malaya to New Guinea. 

YiTTARiA SULCATA, Kubn, Linnssa, xxxvi. 68 ; Hooker & Baker, Syn. Pil. 518. 
At 10,500 feet (Savlland, 1481). 
Distribution : Ceylon ; Perak ; Sarawak ; New Guinea ; Society Islands. 

AcKOSTiCHUM BicrspE, Hookcr, Sp. Pil. v. 271 ; Hooker & Baker, Syn. Fil. 421. 
From 4000 to 8000 feet {Burbidge). 
Distribution : Malaya to Formosa and the Loo-Clioo Arcbipelago ; New Guinea. 

ScHiz^A MALACCANA, Baker, in Hooker & Baker, Syn. Fil. 428. 
At 5000 feet {Burlidge). 
Distribution : Malaya and Philippines 

Angiopteris evecta, Hoflfm. in Comm. Goett. xii. p. 29, t. 5 ; Hooker & Baker, Syn. 
Fil. 440. 

On river-banks in sandstone hills, at 2000 feet [Low). 

Distribution : From Polynesia and Queensland to Japan, the Himalaya, Ceylon, and 
Madagascar. 

LYCOPODIACEJE. 
(By J. G. Baker.) 

Ltcopodium ceylanicum. Spring, Mon. i. 37. 
At 11,000 feet {mivilcmd, 1410). 
Distribution : Ceylon ; Borneo. 

LYCOPODirM SERRATUM, Thunberg, Fl. Jap. 341, t. 38. 
At 4000 feet (Havilcmd, 1416). 
Distribution : From India to Japan and Polynesia ; Mexico ; Bourbon (?). 

Lycopodium Selago, Linn. Spec. Plant, ed. II. 1565. 

At 11,000 feet (Maviland, 1411). 

Distribution : Arctic and temperate zones of both hemispheres ; New Zealand, Australia, 
Antarctic America ; Fast Himalaya. 

Lycopodium Wighttaxi m, Wallich, Cat. n. 2184. 
At 11,000 feet {Uariland, 1412). 
Distribution : Nilgherries and Ceylon ; Java, Borneo. 



DE. O. STAPl' ON THE FLOEA OF MOUNT KIXABALU. 255 

Lycopoditjm clavatum, Linn. Spec. Plant, cd. II. 15G1. 
At 11,000 feet [Hamland, 1413). 
Distribution : Cosmopolitan. 

LYCopODirM CASTJAKiNOiDEs, Spring, Mon. i. 91. 

At 6000 feet (Biirhidge) ; at " lii.i^h altitudes " {Low) ; at 8700 feet (IlaviUmd, 1411). 

Distribution : Malaya to Philippines, and to tlie East Himalaya and China. 

" The indurations on the stem are hard enough to scratch the hands whilst travelling, 
and the trailing stem will, like a rotan, catcli the foot " (Ramland). 

Lycopodium volubile, Porster, Prodr. 86. 

At 8000 feet [Low). 

Distribution: Prom Polynesia, -New Zealand and North Australia, over Malaya to 
Penang and Eorneo. 

SELAGINELLACE^. 
(By J. G. Baker.) 

Selaginella caulescens, Spring, Mon. ii. 158. 

On Daliombang Kiver, at 3000 feet {Ilaviland, 1415) ; common on rocks along the 
streams (Burbidge). 

Distribution : Malaya to New Guinea and Philippines, India to China and Japan. 

Selaginella "Willdenowii, Baker, in Gard. Chron. 1867, 950. 
Prom 4 to 6 feet high, near Kiau {Burbidge). 
Distribution: Malaya; East Himalaya; Cochin-China. 

EQUISETACE^. 
(By J. G. Baker.) 

Eqtjisetum debile, Roxburgh, in Vaucher, Mon. 387. 

On the banks of the Kadamaian River below Kiau, from 2000 to 2500 feet {Low ; 
Ilaviland, 1353). 

Distribution : Tropical Asia to Piji Islands. 



MUSCINE^. 
(By W. MITTE^^ A.L.S., and C. H. Wright, E.R.M.S.) 

DiCRANTJM ASSIMILE, Hampc, Icon. Muse. t. 24. 
At 6000 feet {Saviland, 1428 ; Bow). 
Distribution .- Java, Sumatra, Perak, Sikkim. 

DiCRANUM, sp. aifine -Z>. undulato, Turn. 
At 11,500 feet {Haviland, 1424). 

2l2 



'256 DE. O. STAFF OX THE FLOEA OF MOUNT KINABALU. . 

Campylopus exasperatus, Brid. Bry. Univ. i. 473. 

At 9000 feet (Loiv) ■ Penokok, at 3200 feet (ITaviland, 1421). 
Distribution : Java ; Sumatra ; Sandwich Isl. 

EucAMPTODON EDENTULTTS, Jaeg. Adum. i. 148, var. an&tjsttjs, G. H. Wright, n. var. 
Ab JS. edentulo differt f oliis angustioribus, enerviis. Cellulse apicales cseteris similes, 
non ovales aut laxiores ut in E. macrocalyci, A. Br. [Holomitrium enerve, Dozy & 
Molkenh.). 

Maripari, at 5000 feet (Mavikmd, 1420). 

Distribution (of species) : Ceylon, Borneo. 

Leucobhyum sanctum, Hampe, in Linnaia, xvii. 316. 
Without precise locality (JBurhldge). 
Distribution : Japan ; Malacca ; Java ; Fiji ; Samoa. 

LErcoPHANES SQTJAniiosuM, Brid. Bry. Univ. i. 764. 
At 6000 feet ( 'Burhidge). 
Distribution : Philippines. 

ScHisTOMiTRiUM Lowii, Mitt, in Journ. Linn. Soc. xxii. 302. 
Without precise locality [Low). 

Syrrhopodon involtjtus, Schwagr. Suppl. ii. pt. 1, 117, 1. 132. 
Without precise locality {JBurhldge). 
Distribution : Malacca ; Singapore ; Moluccas ; Samoa, 

Syrrhopodon rueescens, Hook. & Grev. in Brewster's Edinb. Journ. of Sci. iii. 227. 
Without precise locality (Burhidge), 
Distribution : India ; Singapore ; Penang. 

Syrrhopodon alboyaginattts, Schwagr. Suppl. ii. pt. 1, 112, t. 131. 

Without precise locality {Burhidge). 

Distribution : Moluccas ; Piji ; Admiralty Isl. ; Samoa. 

Uhacomitritim jayanicum, Dozy & Molkenb. in Zollinger, Syst. Verzeich. p. 32. 
At 11,500 feet {Ilaviland, 1423). 
Distribution : Java ; Japan. 

Schlotheimia splendida, Mitt., n. sp. Caules repentes cum ramis simplicibus rami- 
ficatisve in csespitem aggregati ; rami fructiferi pollicares, humefacti cum foliis 2\ 
lin. crassi. Polia ramea humefacta squarrosa, exsiccata laxe adpressa, oblongo- 
lanceolata, supra medium semitorta, subcymbiformiter excavata, IJ lin. longa, ad | 
lin. lata, apice acuta obtusave, basi contracta, margine integerrimo subflexuoso, 
Isevia, subnitida, nervo angusto rufo in pilum f lin. longum flexuosum Itevem 



DE. 0. STAFF ON THE PLOEA OF MOUNT KIXABALU. 257 

excurrente, cellulis inferne elongatis, apicera versus sensim abbreviatis, in lineis a 
nervo oblique-ascendentibus dispositis ; folia pcrichaitialia crecta, quam ramea 
angustiora et duplo longiora, pilis longis ercctis seta? medium fere attingcntibus. 
Theca in setfi gracili 5 lin. longA, cylindrica, coUo sensim attonuato I lin. longo. 
operculo basi cupulato, subulatim angustato, sporangio irquilongo, I lin. longo, 
peristomii exterioris dentibus angustis, exsicttindo rcvolutis, humefactis in conum 
conniventibus, peristomii interioris ciliis paulo quam peristomii exteriores dentes 
longioribns erectis; calyptra tliecam totam obtegens, basi biciniata, })ilis brevibus 
adpressis aspersa. 
Maripari, at 5000 feet (Ilarilaud, 1119) ; without precise locality (Low, Burhidge). 
Kemarkable from its thick branches, with le^ivcs not undulated, the cells every wliei'c 
distinct and smooth, and its minutely pilose calyptra. S. Wallisii, C. ]Miiller, Linn;ea, 
1870, p. 173, as described, appears to be very near; but its calyptra is said to bo 
*' dentibus asperis robustis scabra," and there are some other differences. 

ScHLOTHEiMA RUBiGiNOSA, C. H. Wright, n. s]). Folia ligulata, obtusa vel emarginata, 
non pilosa, in medio profvmde uniplicata. Rhizoma repens, necnon apico ia(li(':iu><. 
Rami erecti, 1 poll, alti, simplices, rarius ramosi. Folia imbricata, ligulata, obtusa, 
medio longitudinalitcr profunde unipliccita ; costa in plica^ fundo apice leviter 
excurrens ; cellula3 inferiores oblongae, leptodermatica?, superiorcs in seriebus angulo 
45° costse regulariter dispositoe. Costa foliorum pericha^tialium in pilum laming 
duplo longiorem producta. Fructus in apice ramorum. Seta ] poll, longa ; theca 
cylindrica, basi acuta ; dentes peristomii 10, in medio striatis. Calyptra capsulA, 
longior, cylindrica, basi 4-tida, Isevis. 
At 10,500 feet {Savlland, 1418). 

The aspect of this species is very similar to that of Schlothelmia splendida, Mitt., but 
the leaves are totally diflFerent. Those of S. sjylendlda are acute and have the nerve pro- 
duced into a long pilum, while those of this species are obtuse or even emarginate, with 
the nerve excurrent as a small tooth. The plication of the leaf forms a single pouch-like 
fold, commencing a short distance below^ the apex and continuing nearly to the base of 
the leaf; along its base runs the nerve, which is minutely toothed along the back. "When 
viewed from the dorsal side, this plication resembles the dorsal fin of certain fishes. 

Macromitrium cuspidatum, Hampe, Icon. Muse. t. 20. 
At 11,000 feet (Eaviland, 1417). 
Distribution: Java; Sumatra. 

Beyum PACHYP03IA, Mont. in Ann. Sc. Nat. ser. 3, iv. 104 (1845). 
Without precise locality [Burbidge). 
Distribution : Java. 

Mnilm rostratum, Schwagr. Suppl. i. pt. 2, 13G, t. 79. 
At 5000 feet {Haviland). 
Distribution : Cosmopolitan. 



258 DE. O. STAPP OX THE FLORA OF MOUIS^T KINABALU. 

Uhizogonium distichum, Brid. Bry. Univ. ii. 665. 

Witlioiit precise locality (Burhidge). 

Burbidge's specimens consist of a few barren stems which appear identical Avith New 
Zealand specimens. 

Ehizogonium spinifoume, Brnch, in Flora, 18i6, p. 134. 
Without precise locality {Burhidge). 
Distribution : South tropical and subtropical zones ; also Japan and West Indies. 

Dawsonia supeeba, Grev. in Ann. & Mag. Nat. Hist. xix. (1847), 226, t. 12. 
Without precise locality [Burhidge). 
Distribution : Australia ; Tasmania ; New Zealand. 

PoGONATUM MACROPHYLLTJM, Dozy et Molkcnb., Bry. Jav. i. 45, t. 35. 
At 6000 feet {Loio, Burhidge). 
Distribution : Java, Sumatra. 

Bhacelopus acaijlis, Mitt., n. sp. Caulis brevissimus. Folia erecta, a basi dilatata con- 
tracta, ligulata, apice acuta acutatave, integerrima, cellulis superioribus subrotundis, 
inferioribus oblongis, perichsetialia longiora obtusiuscula, apice paucidentata, omnia 
nervo satis valido percursa. Seta pollicaris, fusco-rubra, papillis asperrima. 
Theca oblonga, demum inclinata, Isevis, infra os contracta, operculo rostrato. 
Calyptra, indumento pallido, thecam totam tegens. 

With old and young fruit (Burhidge). 

Size that of B. pllifer (Dozy et Molkenb., Bry. Jav. i. 37), but foliage entirely different. 

Haerisonia alpiiVa, C. H. Wright, n. sp. H. SumholdticE , Spreng., afEnis, sed folio 
apice non in pilum producto, capsula laevi non sulcata. Caespitosa, brunnea. Caulis 
pinnatim ramosus. Folia dense imbricata, panduriformia, basi et marginibus rubris, 
apice acuminata, subsquarrosa, cellulis vermicularibus, dense aggregatis, margi- 
nalibus elongatis, rubris, translucentibus, alaribus oblongis, rubido-brunneis ; foHa 
perichsetialia oblongo-acuminata, apice hyalina; cellulis elongatis, poris connexis, 
alaribus quadratis. Seta 6 lin. longa, recta ; theca ovalis, laevis, gymnostoina. 
Sporse magme, muricatae. 

On wet rocks, at 11,500 feet {Haviland, 1425). 

Sematophyllum pandtjrieorme, C. H. Wright, n. sp. S. convoluto, Jaeg., affine, sed 
differt statura majore, magis ramosa, foliis sub apice contractis, argute serratis. 
Bhizoma repens. Bami erecti, plus minusve pinnatim ramosi. Folia subpatentia, 
elongato-ponduriformia, 2 lin. longa, margine argute serrato, ex seriebus 3-5 cellu- 
larum elongatarum pachydermaticarum composito, cellulis centralibus minoribus, 
quam marginalia minus crassis, poris communicantibus ut in Dicrano (sect. 
Eudicrano) ; folia perichtetialia 3 lin. longa, late lanceolata, apice in pilum longum 
dentatum producta, cellulis elongatis, poris connexis, marginalibus consimilibus. 
Seta 2-3 poll, longa, rubra, Isevis ; theca inclinata, ovata, 2 lin. longa ; peristomium 



% 



DR. O. STAFF ON THE FLOEA OF :\rorNT KIXABALI7. 259 

exterius ex dentibus 16 transversatim striatis, iiitcrius ex deiitibus 10 dilute flavis 
basi coalitis comj^ositum. Sporae globosoe, dilute flavae. 
At 6500 feet {Hamland, 1426). 

Taxithelium DuBYANrM, Mitt. Ilypmim Duhyanum, C. Miiller, Syn. jMusc. ii. 241 ; 
Dozy & Molkenb. Bry. Jav. ii. 182, t. 280. 
At 5000 feet {Raviland, 1131 ; Burhidge). 
Distribution : Java, Amboina. 

Mniodendron microloma. Mitt., n. s^x Stii:)es liumilis, erectus, pollicaris, rufo-tomen- 
tosus, apiee ramis brevibus, simplicibus, patulis, in comam rotundatam dispositis, 
innovatione conformi continvms, iuterdum iterum iterumque prolifer. Eolia stipitis 
laxa, squaiTOsa, a basi deltoidea subulata, apice denticulata, nervo ad apicera 
percursa ; folia ramea densa, undique patentia, ovato-lanceolata, subulatim 
attenuata, rigidula, e viridi fuscescentia, in dorso et in inarginibus denticulata, 
nervo excurrente, cellulis interioribus angustissimis, marijinalibus minutis, rotundatis, 
limbum indistinctum forniantibus ; folia j)ericlurtialia a basi ovatix pliciitA subulata, 
apice remote denticulata, nervo longe excurrente. Tlieca in seta IJ poll, louijil 
rubra horizontalis, cylindrica, subarcuata, 3 lin. longa, exsiccando striata, coUo 
distincto, operculo magno longirostri, pcristomii dentibus apice angustissime 
attenuatis. 
At 6000 feet {Burhidge). 

Very similar to the Isotheciimi marginatum. Hook. fil. & AVils., figured in the ' Flora 
of New Zealand,' but smaller in all its parts. 21. Korthalsii, Bry. Jav, t. 236, appears 
considerably different in habit, and its leaves have a thickened margin. 

Mniodendeon Koethalsii, V. d. Bosch et Lac. in Dozy & Molkenb. Bry. Jav. ii. 
139, t. 236. 
At 5000 feet {Hamland, 1430). 
Distribution : Java ; Sumatra ; Aneitura. 

MxiODENDROX DiVAUiCATUM, Liudb. in Ofvers. Kongl. Vet. Akad. Porli. (Stockholm) 
xviii. 375 (1861). 
Without precise locality {Low). 
Distribution : Java ; Sumatra ; Malacca. 

Htpxodendron Eeixtvaedti, Dozy & Molkenb. Bry. Jav. ii. 135, t. 233. 
At 6000 feet {Baviland, 1429). 
Distribution: Java; Celebes. 

Hypnodendeox brevifolium. Mitt., n. sp. Stipes pollicaris, basi radiculosus, sparse 
foliatus, ramis pluribus simplicibus vel ramificatis in comam tabuliformem rotun- 
datam 1 poll, latam dispositis. Folia stipitis adpre>sa, subdeltoideo-ovata, in 
margine serrulata, ad medium usque nervo tenui percursa ; folia ramea lateralitei- 



260 DR. 0. STAFF ON THE FLOEA OF MOUNT KINABALU. 

compressa, patentia, exsiccando parum niutata, ovato-lanceolata, in margine serrata, 
nitida, nervo sub apice evanido, cellulis angustissimis elongatis. Perichgetia plura 
aggregata, foliis subulatis remote denticulatis. 

Without precise locality {Burhidge). 

In habit more like a small Mniodendron comosmn, Lindb., but with sinning leaves. 

Rhacopillm spectabile, Reinw. & Hornsch. in Nov. Act. Cses. Leop.-Car. Nat. Cur. 
xiv. pt. 2, 721, t. 40, fig. c. 
At 6000 feet {Biirhidge ; Haviland, 1427). 
Distribution: Java; Sumatra; Philippines." 

AcANTHODirM TMSMEGiSTrM, Mitt. — Syn. Sypnum trismegistum, Mont, in Lond. 
Journ. Bot. 1844, p. 633 ; Hi/j^mim lancifolium, C. Miill. Syn. Muse. ii. 497 ; Dozy 
& Molkenb. Bry. Jav. ii. 141, t. 239 ; Neckera lancifolia, Harv. in Hook. Ic. 
PL t. 21. fig. 5. 
"Without precise locality (Burbidge). 

As in other species of this group, the setic are three inches in height. 
AcantJiodium lancifolium, Mitt., in Harvey's sj)ecimens has complanate panduriforni 
foliage, the apices wide, very unlike Montague's species. Another species, as it appears 
to be, was found in Moulmein by the Rev. D. Parish. There are also a number of 
Indian mosses which appear to belong to the same group ; they are Hypnum 
tanytrlchum, Mont., Stereodon j^sllurus, Mitt., S. surcularis, Mitt., S. pilosulm. Mitt., 
and ;S'. peniGillatus, Mitt., as well as Hypnum papillaUmi, Harv. H. calderense, Sulliv. 
in U.S. Expl. Exp. t. xv., may be a state of A. rigidum, Mitt., and similar specimens 
have been collected in Palawan, in the Isle of Pines, and in Samoa. Hypnum 
exteniiatum, Brid., from Australia and New Zealand, belongs to this group. 



HEPATIC^. 

The following Hepaticae were found growing with the above-mentioned Mosses, but 
several in insufficient quantity for specific determination. The numbers affixed to them 
correspond with those of the Mosses amongst which they were found. 

GOTTSCHEA, Sp. 

At 6000 feet (Bcmland, 1429). 

Plagiochila tenuis, Lindenb. Mon. Plasrioc. 50. 

The fragments found appear to belong to this species, which was previously known 
from the West Indies, Sandwich Islands, and Nepal. 

At 6000 feet {Bamland, 1429). 

Odontoschisma Sphagni, Dum. Rev. Jung. 19. — Syn. Sphagnoccctls communis, Nees. 
This common European species was found at an elevation of 11,500 feet. (Haviland, 
1425). 



DR. 0. STAFF ON THE FLORA OF MOUNT KINABALU. L>(j] 

Chilosctphus aselliformis, Nees, Hepat. Jav. 2G. 
Maripari Spur, at 5000 feet {Haviland, 1419). 
Distribution: Java. 

Saccogtna muricella, Mitt. — Syn. Chiloscyphm muricellus, De Not. Epatiche di Borneo, 
24, t. 16. 
Without precise locality {Burbidge). 
Distribution : Borneo. 

ScAPANiA LEPiDA, Mitt, in Journ. Linn. Soc. v. 101. 
Without precise locality (Burbidge). 
Distribution : Ceylon. 

Lepidozia holorhiza, Nees, in Gott. Syn. Hepat. 210. 
Without precise locality {Burbidge). 
Distribution: Java. 

Lepidozia cladorhiza, Nees, in Gott. Syn. Hepat. 210. 
Without precise locality (Burbidge). 
Distribution : Java 

Lepidozia subintegra, Lindenb. in Gott. Syn. Hepat. 201. 
Without precise locality (Burbidge). 
Distribution: Java. 

Lepidozia trtchodes, Nees, in Gott. Syn. Hepat. 203. 
Without precise locality (Burbidge). 
Distribution : Java ; Bolivia. 

Lepidozia Wallichiana, Lindenb. et Gott. Syn. Hepat. 204. 

At 6000 feet (Haviland, 1428) ; without precise locality (Burbidge). 
Distribution : Nepal. 

Mastigobryum, sp. 

At 6000 feet (Haviland, 1427). 

Trichocolea tomentella, Nees, var. plum a, Nees, in Gott. Syn. Hepat. 237. 
At 6000 feet (Haviland, 1428) ; without precise locality (Burbidge). 

Phragmicoma, sp. 

At 5000 feet (Haviland, 1431). 



second series — BOTANY, VOL. IV. 2 M 



262 



DU. O. STAFF OX THE FLOEA OF MOUNT KINABALU. 



EXPLANATION OF THE PLATES. 
Plate XL 

A. Decaspermum ViUs-Jdaa : — 1. Brancli (nat. size). 2. Flower (X2). 3. Young fruit (x4). 
4. Flower^ vertical section (x8). 5. Petal (xl). 6. Stamen (x20). 7. Ovary^ transverse 
section ( X 20) . 8. Seed (x 20). 9. Seed, longitudinal section ( X 20). 

J^. lEugenia kinahaluensis : — 10. Branch (nat. size). 11. Flower, vertical section (x 5). 12. Petal (x 5). 

C. Eugenia ampullaria : — 13. Fruiting brancli (nat. size). 

D. Tristania elliptic a : — 14. Brancli (nat. size). 15. Flower (x 3). 16. Flower, vertical section ( x 6) . 

Plate XII. 



A. Hedyotis protrusa : — 1. Brancli (nat. size). 

taken off to show the protruding capsule ( 
1>. Hedyotis pulchella : — 4. Branch (nat. size). 

C. Hedyotis macrostegia : — 7. Branch (nat. size) 

and calyx, the front lobes taken off ( X 3) . 

D. Argostemma gracile : — 10. Plant (nat. size). 



2. Flower (x 4). 3. Fruit; the anterior calyx -lobe is 

x6). 

5. Flower ( X 3) . 6. Frnit ( X 3) . 

. 8. Flower, with supporting bracts (x3). 9. Ovary 

11. Corolla (x2). 12. Calyx (x2). 



Plate XIII. 

A. Streblosa urticina : — 1. Branch (nat. size). 2. Branch of inflorescence ( x 6). 3. Flower-bud ( X 20) . 

4. Young corolla, cut open (x 20). 5. Mature stamens (x 40). 6. Very young fruit, vertical 
section (x20). 7. Style (x20). 8. Ripe fruit (x8). 9. Fruit, vertical section (x8). 
10. Fruit, transverse section ( X 8) . 

B. Psychotria densifolia : — 11. Branch (nat. size). 12. Flower (x 8). 13. Pyrene, transverse section 

(X4). 

Plate XIV. 

A. Diplycosia pinifolia : — 1. Branch (nat. size). 2. Flower-bud (x5). 3. Ovary, with disc (xl5). 

B. Diplycosia kinabaluensis : — 4. Branch with adult leaves (nat. size). 5. Flowering branch (nat. size). 

6. Flower (x 2). 

C. Diplycosia pendida : — 7. Branch with young fruits (nat. size) . 

D. Diplycosia riifa : — 8. Branch, one leaf turned, seen from the back (nat. size). 9. Corolla (x2). 

E. Diplycosia chrysothrix : — 10. Branch (nat. size). 11. Flower (x2). 12. Calyx and ovary, vertical 

section (x2). 13. Stamen (x4). 14. Ovary, with disc (x2). 

Plate XY. 

A. Rhododendron lacteum : — 1. Brancli (nat. size). 2. Ovary (x 2). 

B. Rhododendron cuneifolium : — 3. Branch (nat. size). 

C. Gaultheria borneensis : — 4. Branch (nat. size). 5. Flower, vertical section ( x 4). 6. Stamen (x 12). 

D. Vaccinium cordifolium : — 7. Branch (nat. size). 8. Flower (x 2*5). 9. Calyx and ovary, vertical 

section (x2). 10. Stamen (x6). 11. Ovary, transverse section (x 3). 

Plate XVL 

A. Havilandia borneensis: — 1. Branch (nat. size). 2. Leaf (x3). 3. Flower (x6). 4. Diagram 
showing tbe prefloration of the corolla-lobes. 5. Corolla and ovary (x 6). 6. Calyx, final state 
(x3). 7. Ovary (X 12). 8. Nutlet (x 6). 9. Nutlet, transverse section ( x 6) . 



DR. O. STAFF OX THE FLOKA OF MOUNT KIN A HALF. 2G3 

Vk Euphrasia borneensis :~10. Flowering plant (nat. size). 11. Flower (x3). 12. Caly.x (X4). 

13. Calyx with ovary (x 3). ll. Leaf (x 5). 15. Capsule (x 5). 10. Seed (maf^Miified). 
C. Cijrtandm Clarkei :—\7 . Branch (nat. size). 18. Corolla (nat. size). 19. Calyx, enclosed by the 

connate bracts (nat. size). 

Plate XVII. 

Scyphostegia borneensis : — 1. Branch with remainder of an inflorescence (nat. size). 2. InHoreseene(> 
(nat. size). 3. Flower with funnel-shaped bracts and a young accessory bud (x 4). 4. Flower, 
vertical section (xi). 5. Flower-bud ( x 4). 6. Diagram showing the position of the perianth- 
lobes. 7. Half-mature carpel with scales at the base (x5). 8. Very young carpel (x30). 
9. Upper end of carpel, vertical section (xlO). 10. The same, seen from the front (xlO). 
11. Ripe fruit (nat. size). 12. Part of the ripe fruit, carpels with their scaly involucres at the 
base and a portion of the receptacle behind them (x2"5). 13. Ripe carpel, vertical .section 
(XlO). 14. Embryo (xlO). 

Plate XVIII. 

A. Quercus Havilandii: — I. Branch (nat. size). 2. Female flower ( X 4). 3. Segment of the perianth 

with a stamen in front (x5). 4. Female flower, vertical section (x4). 

B. Ficns setiflora: — 5. Fruiting branch (nat. size). G. Receptacle, vertical section (x2). 7. Aclie- 

nium with perianth (xl5). 8. Perianth, segment seen from within (x30), 

C. Loranthus centiflorus \ — 9. Branch (nat, size). 10. Flower (x 2). 

Plate XIX. 

A. Elatostemma buibothrix :—\. Branch (nat. size). 2. Young fruit with supporting bracts (X8). 

3. Diagram of female flower, showing the arrangement of the bracts and staminodes. 4. Stami- 
nodes ( X 20). 

B. Elatostemma thalictroides : — 5. Branch (nat. size). 6. Pedicel of male flower with bract and bract- 

lets ( X 8). 7. Young male flower ( x 8). 8. Perianth of male flower with one stamen ( X 8). 

C. Elatostemma lithoneuron : — 9. Branch of a male plant (nat. size). 10. Male inflorescence (x2). 

11. Male flower in bud (x8). 12. Young male flower, open (x8). 13. Female inflorescence 
(X3). 14. Portion of female inflorescence (xlO). 15. Female flower with supporting bract 
(X20). 16. Fruit (x20). 
1). Pilea Johniana : — 17. Branch (nat. size). 18. Branchlet ( X 4). 19. 4-nier()u.s perianth of a male 
flower (x20). 21. 2-merous male flower (x20). 22. Stamen (x20). 23. Female flower 
(x20). 24. Bracts and staminodes of a female flower (x 20). 25. Fruit (x 10). 26. Fruit, 
transverse section ( x 20). 

Plate XX. 

A. Patersonia borneensis: — 1. Branch (nat. size). 2. Lobe of stigma (X3). 3. Stamens, the tube cut 

open (x3). 4. Ovary, with style and stigma (nat. size). 5. Ripe capsules with supporting 
bracts (nat. size). 6. Capsule with seeds (nat. size). 

B. Patersonia Lowii : — 7. Branch (nat. size). 8. Lobe of stigma ( X 3). 9. Ovary with style and 

stigma (nat. size). 

C. Deyeuxia epileuca: — 10. Plant (nat. size). 11. Spikelet (xlO). 12. First glume (xlO). 

13. Second glume (xlO). 14. Third glume, with palea and produced rhachis (xlO). 
15. Flower (x 20). 16. Caryopsis ( X 10) . 17. Transverse section of leaf ( X 40) . 



TT?ANs.LnNiN,Soc.SER;'Z EoT.VoL , lY. Pi,. II. 




,K,?ita|-i ckL.et libk 



J. N. FitcK irnp 



A.1-9.DECASPERMUM VTTIS IDAEA. B, 10-12. EUGENIA KI NABALUEMSIa. 
*C.13. E. AMPULLARIA, T). 14-17, TRISTAWTA ELLIPTICA . 



Trans.Ltnn.Soc.5er.2Bot.Vol.1VPl.12. 



^X^., 




LH-Fitck del.et \A 



A. 1-3. HEDYOTIS PROTRUSA . B, 4-6 H FULCHELLA . 
G, 7-9. i-L.MACROSTEGIA, D. 10-i^ xac ,MMA GRACILE 



J N. Yifcck iBig. 



Trans, Linn. Soc.Ser.2Bot,Vol. iV Pi ,13. 




C.KRfccK deLeb liik , 



A,MO-, STREBLOSA URTICINA. B. 11-13 , PSJCHOTRTA DET^STFOIJA 



J.¥,FitoK jmp. 



Traks ,Ltnn. Soc . Sttr . 2.Bor?/oL .W^Pl .14*. 




•.!i iif'.eb liiii 



A,1-3,D1PLYC0S1A PINIPOLIA B. 4-6 D. KlNAEALUEl^lSIS C 7,p.PE~NDULlFL0RA 
D.R-9.D.RU1^A E lO-l^.D.CHRYSOTHRIX . 






Trans . Linn . Soc . 8er.2.Bot.Vol . TV.Pl .15 , 

1 .1 i^. 







■fn 








^ X 



^^^. 



V-,^ 



A.l-2.RHODODET^DR0'N LACTEUM . B.3.R. GUl^EirOLIUlvr , 
C.4-6. GAULTHERIA BORKEENSIS. D.7-11.VACC1T^1UM CORDIPOLIUM 



Iv Fitch, imp. 



Trans Linn. Soc.Ser.2Bot.Vot, IV Pl. 16 










7S 



C.H.Piickdel.efclibk. 



73 




A. 1-9. HAVILAWDTA BOI^IMEENSIS . B, 10-16. EUPHRASIA BORNEEI^SIS 
C.17-L9. CYRTANDEA CLARKE I . 



J. N. Fibch imp . 



Tt! A"Ns . L mN .Soc.Se:r.2/Bot Vox, .IV. Tl .17, 




^^Vf I '//■ 



C,HFtbck diilet ]itK, 



J.N.FibcK imp. 



1-14-. SCYPHOSTEGIA BORl^EET^SIS 



■V, "^ <»-«,, 




Trans. Linn, 5oc.Ser.2Bot.Yol IV PlJB 



V ; ^'■;"^ 













C,;H Fitch cl£Leblith 



J.N.Piboli imp. 



A.L-4 QUERCUS HA\ . B.5-8.7ICUS SETIFLORUS 

G. 9-lD. LoKr..Minu5 CENTIFLORUS . 



THANS,LjNN.S0C.SEF.2.BoT.VbL.TV.?L.19. 




C.H.FitctideletlibK. 



A 1-4 ELATOSTEMMA BULBOTHRIX .B.S-8.E THALICTROIDE S 
' C. 9-16. E.LITHONEUROTT.D. 17-26. PILEA JOHNIAl^A . 



J.N, Fitch- ifT-j) 



Tt^aks .LTnh. Soc , Ser. Z.BoT.Vox. ,1V". Pi, , 20. 



CKFtfccKc: 




A 1-6 PATT 



P> PATERSONTA LOWIl 



LINNEAN SOCIETY OF LONDON. 



MEMORANDA COKCERMNG TRANSACTIONS. 

The JKrst Series of the Transactions, contaiTimg both Botanical and Zoological contributions, has been completed 
ia30¥<^.jand a few entire sets are still for sale. Only certain single volumes, or parts to complete sets, may 
^'olil^iii^ at the original prices. The price of the Index to Vols, 1-25 is 8s. to the public, and 6s. to Pellows ; to 
Tols. 1^^-30, 4». to the public, and 3s. to Fellows. 

^fe^^flffid 8^ie» of the Transactions is divided into Zoological and Botanical sections. The prices of the 
SE^PSshI pffte.of t^ese which have been published are as undermentioned. (For the Zoological parts see Zoological 



VolaBie. 






h Sb3^ 1 ism 

xMMm. .... 




Part iS.:i879 



¥m''"M, 1'^ 

0. P«et 1.1^1 

iB^f'H. 188^ ... 

/Jpfe'ttfeBsa ... 



Price to the 
PuWic. 

£ s. (L 

—BpTAITT. 

8 

16 

12 

10 

14 

6 

4 

1 







1 
. 1 
. 1 
. 1 

. 1: 
.050 
. 10 
.030 
. S 
. 13 6 
. & 6 

. i«> 
, Q r a 

.034 




Price to 
£ 9, d. 



6 I 

12 i 

9 I 

7 6 i 

18 i 

19 B ; 

18 ; 

16 ; 

15 

\} \) 

:j 

7 «; 

2 '.'. 

2 3 

10 *i 
7 
7 
5 
2 



Velnme. 



Wlien 
Published. 



Price to the 
Public. 



£ s, d. 
Second Series. — BoT.iwv {irmtinned). 

II. Part XI. 188G 6 

Part XII. 1886. 8 

PartXIIl. 1887 1 7 

Part XIV. 18S7 7 

Part XV. 1887 10 

PartXVI. 1888 2 6 

III. Part I. 1888 3 12 

Part 11. 1891 

Part 111.1891 

Part IV. 1891 

Part V. 1801 



Price to 
Fellows. 

£ s. d. 



Part VI. 1891 

Part VII. 1.^92 

PartVIlI. 1893. . . O 

Part IX. 1893. ... 1 

Part X. 1894 10 

Part XI. 1894 .. . 3 

IV. Part I. 1894 1 

Part II. 1894 2 



4 6 
6 


5 
7 6 
2 
2 14 
3 9 
5 

4 
4 
2 
4 
4 
1 
7 
2 

15 

1 10 



2nd Ser. BOTANY.] 



[VOL. IV. PAKT 3. fe 



THE 



TRANSACTIONS 



OF 



THE LINNEAN SOCIETY OF LONDON. 



THE PHANEROGAMIC ROTANY OF THE 
MATTO GEOSSO EXPEDITION, 1891-92. 



BY 



SFENCEll LE MARCJIANT MOOBIi), B.So., F.L.S,, 



BOTA.VIST TO THE EXPRlttTinX. 




LONDON: 

3V T.AtLaa ASt» WLANCIS. aiSB wow €0C«P. FLSBT '*TSI«rr. 

SOLD AT THE SOCrETt^S APAB'OIBir'm, »tJailSSlX>N*HOI?8»» IPICCABII^T, W,, 

AND JY LON&3£AKS^ SUKEK, A3I1> CO., FATEB:5*0STEm-l^W. 



December 1895. 




MAP TO ILLUSTRATE M" SPENCER MOORE'S 
BOTANICAL EXPLORATION OFTHE 
PROVINCE OF MATTO GROSSO, 1891 



Scalp, 1"?^— 10000^ 



[ 265 ] 



r 



III. The Phanerogamic Boimt^ of the Matto Grosso Eayedition, 1891-92. i>y 
Spencer Le M. Moore, JB.Se., F.L.S., Botanist to the Expedition. 

(Plates XXI.-XXXIX., and Map.) 

llead 2u<i November, 1893. 

Contents. 

Page 

Narrative 265 

Short Sketch of the Climatology of Matto Grosso 277 

The Botanical Geography of Brazil 283 

Flora of Cuyaba and the Chapada Plateau 284 

Flora of Jangada 286 

Flora of Santa Cruz and the Neighbourhood 287 

The Primeval Forest Flora 290 

Flora of Corumba and Coirabni 201 

Kemarks on the probable Origin of the foregoing Floras 293 

Systematic Portion 296 

Description of the Plates 514 

Narrative. 

In the mouth of Juue 1891 I heard from. Mr. Carruthers, Keeper of the Botanical 
Department of the British Museum, that a botanist was required to join an expedition 
about to explore the Brazilian province of Matto Grosso. As I had been for years 
desirous of rendering my botanical education less incomplete by a visit to the Tropics, I 
readily acquiesced in Mr. Carruthers's suggestion that I should offer my services in the 
above capacity. On enquiry it was understood that a syndicate had obtained mineral 
rights over two large tracts of land in the province, the conditions comprising a survey 
and a geological and botanical examination of the territories in question, which included 
the country to the east of Cuyaba, the capital, and the large and almost entirely 
unknown region lying west of the fifty-seventh meridian, and between the tenth and 
tifteenth parallels of latitude. My offer was accepted on the part of the syndicate, and 
on July 2nd, after hasty preparation, I left Southampton in the company of Dr. John 
William Evans, who had been appointed geologist to the expedition. 

Matto Grosso has been visited by several botanists, and, thanks to their united labours, 
the vegetation of a considerable part of the province is now fairly well known. The van 
appears to have been led by a Portuguese doctor, Alexandre Rodriguez Ferreira, who, 
departing from Para towards the end of last century, made his way up the Madeira and 
Guapore rivers to the then flourishing city of Matto Grosso (Villa Bella). He returned 
by the same route, and deposited his plants in the Museum at Lisbon, whence a set was, 
a few years ago, sent to the Herbarium at Kew. In 1825 Riedel, botanist to the Russian 

SECOND SERIES.— BOTANY, VOL. IV. 2n 



266 ME. SPENCER LE M. MOOEE— PHANEEOGAMIC BOTANT 

expedition under Langsdorff, after collecting in EasternBrazil, crossed the Parana river, and 
entered Matto Grosso from the south. Riedel travelled up the Rio Pardo to Camapuan, 
and proceeded via the Taquari, Paraguay, Sao Lorenzo, and Cuyaha rivers to Cuyabji. 
Unaccompanied by Langsdorff, who selected the Arinos and Tapajos route, he advanced 
to Villa Maria and Matto Grosso city, thence descending the Guapore and Madeira 
rivers to the Amazon. Then we find Dr. Patricio da Silva Manso, a Cuyaba physician 
who flourished duriag the earlier half of the century, sending to Lhotsky many plants 
from Cuyaba and the neighbourhood. These plants came into possession of Von Martius, 
and were distributed to various herbaria. Gaudicliaud entered Matto Grosso during the 
second of his three voyages, that of the Hermiaie (1830-33). I have been unable 
to get any information about this voyage, of which no narrative, it would appear, was 
ever written — a singular fact, seeing that the other two journeys were so well and so 
exhaustively described. Por this reason, it is impossible to say to which part of the 
province Gaudicliaud went. Only a small district of Matto Grosso was visited by 
D'Orbigny, viz. the neighbourhood of the Porte do Principe de Beira on the Guapore, 
whence he brought home a few plants. Ten years after Gaudichaud, Dr. AVeddell began 
those travels which have made his name so famous. Pinding himself at Goyaz, Weddell 
moved northward along the Araguaya to its junction with the Tocantins, which river he 
ascended, and, returning to Goyaz, entered Matto Grosso from the east and proceeded 
over the plateau to Cuyaba. Here he turned northward, reached Diamantino, and made 
a short circuit to the Arinos valley and back. Retracing his steps to Cuyaba, Weddell 
went down the Cuyaba, Sao Lorenzo, and Paraguay rivers as far as Olympo ; whereupon, 
being refused admission into Paraguay, he turned back, visited Miranda on the Mondego 
river, and ascended the Paraguay as far as Villa Maria. A journey to Cuyaba and back 
preluded his advance over the watershed to IMatto Grosso city, and from there he turned 
southward into Bolivia. Weddell's valuable collections are at the Paris Museum, and 
comparatively few of his plants have been taken up in the ' Plora Brasiliensis ' of Von 
Martius. Pinally the name of Tamberlik now and again occurs in the work just named, 
as that of a collector in " VTestern Brazil," but of the date as of the locality of his 
explorations I am in complete ignorance, and so cannot even say whether he was in 
Matto Grosso at all. 

Arriving at Buenos Ayres we, on July 28tli, transhipped into the Brazilian mail-boat, and 
proceeded up the River Plate. Our journey was uneventful, and on the morning of 
August 5th the Rio Apa was reached and we entered Matto Grosso. The most 
remarkable feature of this part of the river is the occurrence of forests composed 
exclusively of the Caranda Palm {^Coj)ernicia cerifera^ Mart.). Extending for a 
considerable distance along the shore, these forests reach northward as far as the 
neighbourhood of Coimbra. The massing of these Palms has a very peculiar effect, the 
grey stems simulating fog, from which the feathery crowns emerge clearly into view. 
Wherever a forest fire has raged, the base of their stem is blackened and the low est leaves 
are apt to lose colour and hang pendent ; otherwise fire has no apparent effect upon 
them. The vegetation bordering this part of the river appears, so far as one can jud^-e, 
to be composed largely of Cassia and Mimosa scrub, with Malvaceae, trailing Ipomoeas, 



/ 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 267 

Calystegia sepmm, &c. The large, handsome, upright Ipomaea fistulosa, Mart., is also a 
prominent feature in the landscape, and the Lepacho tree [Tahrhtda Jrrlhtnedofy Tiorentz), 
leafless, and covered, at this season, Avith a cloud of pink blossom, affords a line spectacle 
to the traveller. 

We steamed into Corumbil early in the morning of the 7th. Here it is necessary to 
tranship into a smaller vessel, and as this had not yet arrived at Corumba on its down- 
Avard journey, we were detained a day at that i)lace, the fiscal capital of the province, and, 
as such, a town of some importance. Advantage was taken of this delay, and we went 
for a ramble through the woods, but with disap})ointing results, as, owing to its being the 
middle of the dry season, but fcAv flowers were to be seen and most of the trees and 
shrubs w ere denuded of their leaves. A Stephanophysum * w^as common, and a lovely 
purple-flowered Unellia ; I also noted a JFaltheria (probably W. viscosisdma, A. St.-llil.), 
a small Croton, Momordica Charantia, Linn,, and the fine shrub Cordia iusignis, Cham., 
then in full floAver, while by the inundated bank of the river were E'whornea azurea^ 
Kunth, and Ipomoea fistulosa^ Mart., accompanied in places where the cunciil was not 
felt by Azolla americana in plenty. 

Next morning our boat left for Ciiyaba, Av^hlch city was reached on the 13th. "While 
taking in Avood at the Aarious log-cutting settlements en rouie a fcAV small specimens 
AA'cre hastily gathered ; of these some have jn'oved to belong to new species. An 
exceedingly common feature of the riverside flora is the prevalence of the Avillow-like 
Alchoniea castanoplia, A. Juss., which forms a close fringe to the banks for a long 
distance. 

On the 21st, favourable chances having been meanwhile seized of collecting in tlie 
neighbourhood of the city, a small expedition set off, under the charge of Mr. Arthur 
AVldtehorn, a gentleman then resident in Paraguay, for the Chapada jilateau, lying several 
leagues to the east of Cuyaba, and the westernmost extension of the great plateau of Matto 
Grosso. The country round Cuyaba consists of gently rolling land, fairly Avell timbered, 
and with abundance of shrubby vegetation. The subsoil is a red argillaceous earth, 
during the dry season ])aked as hard as a brick ; this rests upon a platform of ancient 
slates, Avell exposed on the sides of deep Avatercourses Avhich are dry at this time of the 
year. The traveller is struck by the curious appearance of the ground hereabout ; this 
is due to the fact that, for miles around the city, the red earth has been turned over 
by gold-seekers, large quantities of the precious metal having been extracted here 
during the latter half of last century. As Ave passed, the Avoodlands Avere rather desolate, 
much of the vegetation being parched by the drought. Among the trees was seen a 
Cochlospermum (probably C. insigne, A. St.-Hil.), leafless nOAv, but adorned with yellow 
blossom and cottony capsule ; a Eignoniacea, most likely Tabehum aurea, Benth. & Hook, 
f,, bearing masses of golden flowers, and the Vochysiaceous Saherlla concallariodora^ 
A. St.-Hil., its coarse ragged branches crowned with large Avoody fruits. Of shrubs one 
may mention the very common BauJduia ohtusata, Vog., with thick leaves, Avhite petals, 
and lanky pods ; Composites such as Vernonia obovato. Less., and V. scabra, Less. ; 
Malpighiacese [Heteropteris nudicaidls, 8. Moore; Tetrapteris pvcecox, S. Moore; 

* ily presses being packed up, no specimens were collected during this ramble. 

2n2 



268 ME. SPENCER LE M. MOORE— PHANEEOGAMIC BOTANY 

Banisteria constricta, Griseb., an Amazonian plant) ; Cassia dysophylla, Bentli, ; Myrtles 
{31yrcia dasyblasta, Berg, Eugenia miniata, S. Moore) ; the pretty Lippia lasiocalycim, 
Cham., and many others. Grassy vegetation is abundant wherever circumstances favour 
its obtaining a foothold. Arthropogon villosus, Nees, Andropogon (condensatus, H.B.K., 
ternatus, Nees, virginicus, linn.), Melinis minutiflora, Beauv. (the well-known *' capim 
gordura" of the Brazilians), Faspalum ti^opicum, Doell, and the local Fanicum ansatum, 
Trin., are all to be found here. On stream-sides you will see Jussicea nervosa, Poir.; and 
although, with the exception of grasses, lowly vegetation is not abundant in August, yet 
such does occur ; Jiisticia metallicorum, S. Moore, Buehnera picilustris, Spreng., and the 
terrestrial orchid Stenorhynchiis australis, Lindl., may be cited in support of the statement. 
Through romantic scenery you ascend the plateau by a fairly good road, a creditable 
piece of engineering in so remote a part of the republic. Dr. Fonseca*, judging from the 
supposed heights of various places up the river from Buenos Ayres and Rosario to 
Corumba (which place he considers to be nearly 400 feet above the sea), puts the 
elevation of Cuyaba at about 500 feet. This is at the port f, above which the ground 
whereon the city is built rises at least 150 feet higher. There is no sensible increase 
of level until the entrance to the pass is reached ; approaching this, the escarpment of 
the plateau is seen extending on either hand far as the eye can reach, and rising to about 
1800 feet above the lower country, this giving a total elevation of about 2500 feet above 
sea-level. It must be remembered, however, that this estimate, although probably fairly 
correct, is approximate only, the country never having been levelled. AX the entrance to, or 
in, the pass I gathered Erythroxylon daphnites, Mart. ; Selicteres chapadensis, S. Moore ; 
the milk-white trusses of Serjania perulacea, Eadlk., a beautiful climber; Jlirtella 
amerieana, Aubl., and H. collina, S. Moore ; the tall Justicia oreadum^ S. Moore ; Piper 
asperipUum, Buiz & Pav. ; Sjnranthes grandiflora^ Lindl. ; Benealmia foliosa, S. Moore, 
with leaves a metre long, scarlet scapes and purple leaf-sheaths, as well as other plants. 

The plateau is a good instance of that type of country called by Brazilians " cerrado." 

It consists of neglected -looking land with small, copiously branching trees, a certain 

amount of shrubby undergrowth, and grass in plenty, the latter sometimes forming small, 

almost treeless " campos." It is not without justice that travellers have compared these 

cerrados with orchard-land in our own country. Somewhat deep valleys, well timbered, 

and of which the bottoms are the beds of streams, often dry in August, cut into the 

plateau, whose uppermost portion is composed of brick-red sandstone, which weathers in 

places into grey castellated forms of wonderful variety. Among the trees on the plateau 

may be noticed Byrsonima verbasci folia, Eich., with a plentiful show of saifron flowers ; the 

Lythraceous Bhysocalymna scaberrima, Pohl, its cloud of purple blossoms affording a 

splendid sight; Ltihea sp^(?zos<:«, Willd. ; Euphorbiacese such as Map?^ouma guianemis, Aubl., 

and Mahea jistidifera. Mart., Connarus fulvus, Planch., and many others. You are still 

accompanied by Tahebtiia aurea ?, Benth. & Hook. f. ; this you can distinguish even to the 

far distance by a golden gleam amidst surrounding boscage. The grassy plots are now and 

* ' Tiagem ao redor do Brasil (1875-1878),' Rio, 1881, (p. 35). 

t Derby (' Physical Geography and Geology of Brazil,' p. 11, translated from ' Brasil geographico e historico,' of 
J. E. Wappaeus) gives 123 metres (about 400 feet) as tbe height of Cuyaba port. 




OF THE MATTO GKOSSO EXPEDITIOX, 1891-92. 269 

again lit up by the carnation flowers of Pavonia Bosa-sinen^k, A. St.-Hil, and in its neigh- 
bourhood you may gather sweet-scented Kiclmeyera [K. amplexwauUs, S. Moore), Tnrtiera 
chrysodoxa, S. Moore, tufted yEschynomene orobotcfes, Benth., and the straggling \fauihot 
tripartita, Muell. Arg., var. vestita. The Labiate Hyptis cff'ttsa, S. Moore, r(Mu:irkabIe for 
its loDg, capillary, almost invisible peduncles, rises to a man's height. Atelastomacese 
are common — Miconias, Clidemia nibt^a, Mart., Ibcoca fornicaria Mart, the gorgeous 
Tbbonchina stenocarpa, Cogn., &c. ; Lippia laslocalycina, Cham., Chaptalia integrifolia. 
Baker, Trivis ophiorrMza, Gardn. (here, as in Eastern Brazil, a reputed alexipharniic), the 
fine heads of Chuquiraga chapadensis, S. Moore, the lowly Cuphea micmntha, 11. B. K., 
Ewpatorium megaphyllum. Baker, and masses of the common weed Stachytarpheta 
dichotoma, Vahl, will also be met with. Nor will a yelloAv-flowered R-ubiacea (Chonielia 
ribesiodes, Benth.) escape notice, if only from the peculiar habit shown by its inflor- 
escences of sheltering themselves from the broiling sun beneath a neighbouring leaf. 
The grasses, much the same as those met with in the low country, are frequently burnt 
by campo fires, but they suffer no permanent injury. Indeed, specimens l)elonging to 
this order, with charred outer leaf-sheaths, are common in herbaria, and most of my 
Chapada grasses show unequivocal signs of the action of fire. 

V\^e reached the decayed village of Sant' Anna da Chapada on the 25th and, the weather 
being superb, proceeded to sling our hammocks in a wood hard by. A little mountain 
stream, one of the feeders of Coxipo river, flows past the village, and in wet ground on its 
margin Eleocharis sulcata and.£'. ochreata, Nees, Rynchospora glaiica^ Vahl, and Mayaca 
Selloiciana, Kunth, were found, while Eragrostis Vahlii, Nees, and E. multipeSy S. Moore, 
grew in plenty upon a bank skirting the stream. Other grasses of the place are Paspalum 
panicidatum, Linn., and P. coujugatum. Berg., Panicum horizontale^ G. Mey., and Jlelinis 
miniitijlora, Beauv. In shady places Ichnosipha coticinnus, S. Moore, is to be seen, and 
you may collect Xylopia grandijlora, A. St.-Hil., Myrtaceae {Myrcia ambigua, DC, and 
M. chapadensis, S. Moore), Miconias, a tree-fern, Psychotria tomentosa, Muell. Arg., &c., in 
the course of a short ramble. Noteworthy, too, is Panisteria constricta, Griseb., with 
its remarkable dimorphism. This shrub usually branches somewhat sparingly, and the 
branches bear large leaves ; in places, however, the branches are much more numerous, 
and the leaves upon these are very small. We had, unfortunately, but two pack-mules, 
so that my supply of paper was soon exhausted, otherwise a larger collection could have 
been made upon this interesting plateau. We returned to Cuyabd on August 30th. 

I continued collecting at Cuyaba, and, although Manso appears to have secured a 
considerable number of specimens here, was not unsuccessful in finding new species. 
The small tree Anona Walkeri, S. Moore, growing on the city outskirts, and Eupatorittm 
cuyabense, S. Moore, may be cited in instance. Here I found Copaifera elUptica, Mart., 
the Apocynacea, Anisolobus Zuccarinianus, Mart., AUbertia myrciifolia, K. Schum. (a 
common shrub), Bowdichia vlrgilioides, H. B. K., Hyptis {glauca, A. St.-Hil., crenata, 
Pohl, brunnescens, Pohl), and the elegant little Auratea Biedeliana, Engl., rare in 
herbaria, but here abundant. 

At Cuyaba we were joined by Lieut. Olaf Storm, of the Argentine Navy, prospective 
leader of the expedition, and by his brother Mr. John Storm, second in command, and it 



270 MK. SPEXCEE LE M. MOORE — PHANEROGAMIC BOTANY 

being deemed advisable that a party in advance should make its apj^earance, as soon as 
possible, within the district embraced by the Western Concession, on September 7, after 
an interview with the Governor of the Province, we set off under the leadership of Mr. John 
Storm, for Santa Cruz (Barra dos Bugres), situated on the Paraguay river, about 100 miles 
to the north-west of the city. Lieut. Storm himself proposed returning to Corumba, 
where he expected to receive from Europe the necessary funds. There he was to engage 
men, and transfer to the Villa Maria mail-boat the stores already at Corumba. Meanwhile 
a small stern-wheeler, the Explorer, was on its way to Villa Maria uader the charge 
of Mr. Noel Lofft Ilolden, who had directions to ascend the Paraguay and join us at 
Santa Cruz. Thence some of the party were to return at once by the Explorer, the 
rest proceeding with the mules by a difficult track to the Sipotuba river, on the banks of 
wliich a junction was to be effected. As we took but three pack-mules, the bulk of the 
scientific apparatus had to be sent back to Corumba, a circumstance which, although 
absolutely essential, was most unfortunate, since our stay at Santa Cruz lasted much 
longer than had seemed at all likely when we set out. 

The country between Cuyaba and the Paraguay river resembles in its external features 
that which we had already passed through to the east of the city, but it differs in being 
traversed by limestone and sandstone bills having a N.E. and S.W. trend. One of 
these hills, " Arara " by name, is probably nearly 2000 feet high. But before we reached 
the hills, on our arrival at the little settlement of Jangada, the temporary i]lness o^ one of 
the party brought us to a halt, which I turned to account by collecting specimens. The 
beautiful Vochysiaceous tree Callisthene fascicnlata. Mart., was here in full bloom, as also 
was its lowly ally Camarea ericoides, A. St.-Hil. Here, too, were growing the dwarf 
Palm JDiplothemium jangadense, S. Moore, Byrsonima cydoniafolia, A. Juss., Tocoyena 
hirsuta, Moric, with large sweet-scented Gardenia-\\kQ flowers, the tiny Fectis 
jangadensis, S. Moore, most nearly allied to two Mexican species, Fip)er tuberculatum, 
Jacq., Julocroton humilis, Didr., a lanky, nearly leafless Eranthemum, witli blue flowers 
{E. congestuni, S. Moore), the pretty Zephyrmithes lactea, S. Moore, &c. In front of 
the principal inhabitant's house were a couple of large Taruma trees ( Vitex cymosa, 
Bert.), almost leafless at this season of the year, but with a j)lentiful show of Oxford 
blue blossoms. These trees were festooned with the remarkable Loranthad StruthantJms 
polyanthus, Mart., var. mattogrossensis, which was in full flower at the time of our visit. 
Melastomaceae I saw nothing of, a rather curious circumstance. 

It is necessary to ford several streams on the journey to Santa Cruz. These streams, 
many of which are running throughout the year, flow in the bottom of shallow, well- 
wooded valleys. In the case of one of them, the rapid Curupira, the track runs for some 
distance through its valley, so that a person might fancy himself to be in a great forest 
did he not know the contrary. The reason for these narrow riverain mattos is undoubtedly 
that given by Liais *, who holds that the water needful to the growth of the trees is 
derived, while the dry season lasts, from the thick morning mists caused by excess in the 
temperature of the water over that of the air during the early morning hours, 

* ' Climat, Geologie, Fauna et Geog. Bot. du Bresil,' p. 597. 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 271 

Leaving Jangada on the 13th, we crossed the hilly country and at noon on the 15th 
dismounted before a " fazenda " on the Jocoara river, a tributary of tlie Paraguay. Upon 
the banks of this river grows a plant, presumably a species o( Ilr.r, which furnishes a m(if(' 
and is largely used l)y the inhabitants of the district. At the fazenda they were drying 
leafy branches of the shrub in a store, and when dry pounding them with strong 
wooden rods in a deep vessel made of Avood. I did not secure specimens of the plant at 
that time ; but soon after our arrival at Santa Cruz I instructed a messenger (Voiii 
Cuyaba, on his return, to beg the people to send me some material, and a small party 
from the Explorer, wdio went into the city on business some weeks after, also pre- 
ferred the same request, and compliance Avas promised ; for some reason, liowevor, no 
specimen ever reached me. This mate is somewhat darker in colour than the Paraguayan 
kind, and yields a browner liquor, which is decidedly bitter to tlie taste. On September 
16th we saw the waters of the Paraguay flashing in the morning sun, and our journey 
was at an end. 

Santa Cruz, better known in the neighbourhood by the name of Barra dos Bugres, is 
situated on a low bluff upon the further, or, since its course is here almost due east and 
west, northern side of the Paraguay. On the southern side of the river there is here 
much low-lying open campo, and the river is approached tln-ough Bromeliaceous scrub. 
The bluff is composed of exceedingly fine-grained red shah^s, often greasy to the touch — 
these Dr. Evans distinguishes as Matto shales. The ground rises rapidly from tlie river, 
above which the upper part of the settlement stands some 100 feet ; it is here of cerrado 
nature, but immediately beyond you suddenly enter a dense virgin forest, extending 
many miles to the northward. Close by, the small Rio dos Bugres joins the Paraguay, 
and this gives the place its alternative name. The Bio dos Bugres flows northward to 
the hills forming the southern escarpment of the great Amazonian plateau; it receives in 
its course the tributary Bio Brasinho, of which the feeders traverse the forest just 
referred to. 

Of trees at Santa Cruz there are plenty, but the bulk of the vegetation consists of tall 
evergreen shrubs, usually more or less deprived of leaves in their lower part. Agreeably 
with what one sees in the country to the eastward, the trees for the most part do not grow 
very high, and they often have stout, greatly divaricated branches, invested in thick, 
sometimes locally swollen " bark," in all probability functioning as a preventive of 
evaporation. In a ramble round the settlement you will see the Vochysiaceous 
CalUsthene fasclciiUtta, Mart., and species of Qualea {grandiflora,M.'aiVi., imrvijlora , Mart., 
pilosa, Warm.), Bheedia Giiacopanj, S. Moore, with pleasantly acid yellow fruits used 
in place of vinegar by the people and esteemed as a bait for certain fish, and the Stercu- 
liaceous Guazuma ulmifoUa, Lam., much like a Lime in general appearance, its sweet- 
scented flow^ers the resort of neighbouring bees. For the fruit of the tall Siputa 
tree {Salacia Sljyuta, S. Moore) you must go to the river-banks, where will also be 
found the Caja {SponcUas lutea, Linn.), Broslmopsis lactescens, the type of a new 
genus of Artocarpese, also the myrmecophilous TriplcuHs formicosa, S. Moore. 
Jacaranda cuspidifolia , Mart., a species confined to the province, flaunts its dark blue 



272 ME. SPENCER LE M. MOORE— PHANEROGAMIC BOTANY 

blossoms in the breeze ; nor should Curatella a/niericana, Linn., Vismia japurensis^ 
Reichardt, and the Lepacho * be omitted. 

Among shrubs Anonacese are conspicuous, especially Anona coriacea, Mart., var. 
dioica, A. St.-Hil., and Stormia brasiliensis, S. Moore. Species of Selicteres, including the 
curious H. ortJiotheca, S. Moore, with large straight capsule-valves, are common, as are 
the Erythroxylons, especially E. nitidum, Spreng. Bauhinias {Eauhinia obtusafa, Vog., 
£. cumanensis, H. B. K., and others) frequently occur, and Malpighiacese (Heteropteris, 
Banisteria, Byrsonima), Myrtacese {Eugenia, Myrcia, Psidium Aragu^ Baddi, yielding a 
pleasant fruit), Melastomaceae (Miconia, Clidemia, &c.), and Coccolobas, the red-plumed 
tiny flower-heads of CalUandra pa7^viflora, Benth., Heist eria rubricalyx, S. Moore, the 
tall Oxalis hii^sutissima, Zucc, Turnera odorata, Rich., Siparuna guianensis, Aubl., 
Nua he^^maphrodita, S. Moore, Sorocea grandifolia, S. Moore, may be mentioned among 
many others, as well as the East Brazilian Brosimum Gaudichaudii, Tree. Palms, such as 
the Urubamba, the Tocum, the Bocaiiba, the Guacory, occur here and there f. Climbers 
are plentiful : Hirceas {volubilis, S. Moore, and sepium, A. Juss.), Passion-flowers 
{Passifiora tricusjns, Mast., P. coccinea, Aubl.), Bignonias, Dalechampia cuiabensis, Muell. 
Arg., looking like a Bougainvillea with its pink involucral bracts, the fine red-flowered 
Anguria gloriosa, S. Moore, and Smilax medlcinalis, S. Moore, are worthy of notice. 
Another common order is that of the Rubiacese. You will often find little sandy stretches 
by the river-side brightened by the small pink-blossomed gregarious Sipanea veris, S. Moore, 
and mdiardsonia grandijiora, Cham. & Schlecht., and B.pilosa, H. B. K. ; Borreria cupn,- 
laris, DC., Diodia sapotim^ioides, Presl, and D. multiflora, DC, as well as Sabicea 
humilis, S. Moore, are all lowly members of this order, belonging to which one may cite, 
among the shrubs, Coussareas, Faramea coussarioides, S. Moore, Budgea viburniodes, 
Benth., Ifapoiirea tomentella, S. Moore, Psychotria subci^ocea, Muell. Arg., resplendent with 
coralline peduncles and yellow flowers, Chiococca brachiata, Ruiz & Pav., the Amazonian 
Bandia Buiziana, DC, var. longiflora^ K. Schum., and the remarkable new Alibertia ver- 
rucosa, mth warted berries, made into a preserve by the Santa- Cruzans. Among mono- 
cotyledonous herbs you will see, nestling in the shade of shrubby covert, the delicate 
trumpet-shaped golden flowers of Costus acaulis, S. Moore, Calatheas, the new Maranta 
longiscapa, and Benealmia Holdeni, Zygella graminea, S. Moore, ground aroids ( Caladium 
heteroty^icwm and Aphyllarium tid)erosum, both new to science), Herreria Salsaparilha, 
Mart., and several others. A small brook, nearly dry in October, runs through the 
settlement. On its bank or growing in its bed you may come across, inter alia, Aciotis 
dichotofna, Cogn., and Aclsanthora inundata, Triana, both Melastomaceai ; Heliotropes 
[Seliotropium indicum, L., inundatum, Sw., filiforme, H. B. K.), Conobea scrophularioides, 
Benth., a little blue-flowered Scrophularia, and Polygonum acre, H. B. K. Peperomia 

* I did not see this tree in flower at Santa Cruz, but our woodcutters by prefercuce selected dead Lepacho wood 
all the way up the river. If the Santa Cruz plant be not true Lepacho, it is most probably closely allied to it ; at any 
rate, it has the same peculiar greenish wood. 

t I did not pay special attention to this group, and, judging from the few specimens brought back, the various 
local names are here applied to Palms specifically distinct from those known by the same names in the eastern 
pi evinces. 



OF THE MATTO GROSSO EXPEDITIOX, 1891-92. 



•j:i,i 



mimmularifolia, H. B. K., clings to the shady trunks of trees, wherefroni hang the 
gorgeous trusses of Cattleya superha, Schoiub., the liithcn-to nearest known habitat 
of which lies hundreds of miles to the north. 

In grassy places one can gather Paspalums, Fanicums, Setarias, Elemiuc tifJica, 
Gaertn., Leptoohloa domingensis, Trin., and the little new Luziola pntiilla, with wliich 
specimens absolutely identical from Guiana are in the Herbarium at Kew. One may 
complete the list by citing the names of a ^e\\ lowly dicotyledons, such as Riiella hnmUis, 
Pohl, with milk-white flowers, Stenandriums, the exceedingly common Seopuria dnlcis, 
Linn., Blanthera pohjgaloldes, S. Moore, the peculiar Amazonian Sfilpnopappus I'iridis, 
Benth., Bhodocahjx rotund if oil us, M-neW. Arg., and the T^Teiiy \iii\e Desdemon^i pnlchella, 
the type of a new genus of Scrophulariaceae. 

The Bxplorer not having yet succeeded in surmounting the difficulties of navigation 
between Villa Maria and Santa Cruz, an expedition was arranged with the object of 
making its way through the forest to the Serra de Tapirapuan, the geology of which, it 
was thougJit, might prove interesting. The forest covers a tract of land lying some 
150 feet above the Paraguay. At intervals small streams, most of wliicli run eastward 
into the Brasinho, traverse the line of route, which latter soon becomes a narrow foot-track 
used by the " poayeros," or gatherers of ipecacuanha. There are no inhabitants except 
during the period of securing the valual)le root, which, in this jjart of Brazil, is the wet 
season. For the accommodation of the poayeros small groups of palm-thatched huts have 
been erected at various points in the forest. As will readily be understood, the obstacles 
to the passage of a mule-train along such a track, ever liable to obliteration by the 
encroachment of vegetation, are by no means slight, and it is owing 2:)artly to this that 
my collection from the forest-region is not a large one ; it was also necessary to hurry 
on our way, as our supply of food was limited. 

To judge from the descriptions of travellers, this forest has many points of resemblance 
with other primeval forests of the country. Trees, supjiorted sometimes on gigantic 
boles, raise their lofty crowns far above the traveller ; lianes of various kinds everj- 
where interpose to entangle the steps of his mule ; tree-aroids may be seen perched far 
beyond his reach ; epiphytic orchids abound. Except for the roar of an occasional storm 
as it sweeps by, the deepest silence prevails — silence broken only now and again by the 
crick-crick of a Cicada. Shafts of light slant through the living canopy to illuminate 
the metallic lines of great 3forpho butterflies heavily flopping through the semi-darkness. 
Before you entered the forest, the probability is that the intense heat of the sun was 
mitigated by a fresh breeze, and although the sun's rays cannot penetrate the sylvan 
recesses, the wind also is shut out, and you experience a feeling of oppression in 
consequence. Everywhere you will find the same kinds of shrubby vegetation, vegeta- 

I tion specially adapted to flourish in deep shade. Eoremost among these shade-loving 

plants are Rubiaceie, especially species of Psychotria ; the epiphytic aroids are xVnthuriums 
and Monsteras ; Peppers are conspicuous from the curious way in which their flower- 
spikes stand out rigidly from their leaves, as if movements of the latter against the flowers 

I might cause loss of pollen ; the little Peperomia nummular if olia, H. B. K., is everywhere;, 

tall Solanums, Tabern(Bmontana bushes bearing white flowers exhaling a strong scent as 

I SECOND SERIES. BOTANY, VOL. lY. 2 O 



"274. ME. SPENCEE LE M. MOORE— PHANEROGAMIC EOT ANT 

of new-mown hay, and species of JErythroxylon, Alloplectus, Bignonia, IscJmosiphon, 
Dalechampla, Giiarea, Allophjllus^ sw^ell the list of these dwellers in the twilight glades. 
Orchids you may gather in plenty, most of these, however, not in flower during 
September ; chief among those then in bloom are Cattleya superba, Schomb., and Rodri- 
guezia secunda, Ruiz & Pav. (the latter, like the former, hitherto not found in this part of 
Brazil), also Dichseas, Notylias, &c. The chief feature to notice about this forest-flora is 
its strongly pronounced affinity with that of the great Amazonian country to the north. 
I shall return to this matter later on. 

Now and again the track will emerge from dense forest to cross a short stretch of 
■campo, at the most not more than a few hundred yards in width. The origin of these 
small campos must, I think, be ascribed to fires, for it is difficult to understand how 
•coDditions favourable to forest-growth should suddenly cease and become operative again 
at so short a distance. There is, of course, the alternative suggestion that the district 
has been the theatre of much physical change, and that these strips of campo occupy 
the sites of large rivers which formerly drained the region. But there are reasons for 
preferring the former view. Close to Tapirapuan, however, the case is different. There 
the forest ends suddenly, and after passing a narrow fringe of cerrado, you descend upon 
a grassy campo about a mile in width, beyond which, flanked by a similar cerrado fringe, 
suddenly rises the tree-clad Serra de Tapirapuan, here trending almost due east and west. 
One is led to suspect that the waters of a broad river may at one time have flowed down 
this grassy valley, and to-day a small stream, probably larger during the rainy season, 
does actually meander through it. I obtained but few plants on the campo, JBletia catenu- 
lata, Huiz & Pav., a beautiful ground orchid, hitherto found, unless I mistake, only in 
Guiana and Peru, the new Herpestis parvula, a tiny inhabitant of the streamlet, and 
the Cycad, Zamia Brongniartii, Wcdd., being almost all. This Cycad I afterwards 
found in plenty at Santa Cruz, which place marks the S.E. limit of known Cycad 
distribution in South America. 

Our provisions being nearly exhausted, we hastened back to Santa Cruz, from which 
place we had been absent a week, and a few days after, on Oct. 5th, we were joined by 
the party in the Exploiter. With the boat two expeditions were made : in the first of 
these we ascended the Rio dos Bugres, which w^as soon found to be no longer navigable, 
whereupon w^e proceeded in a canoe up that river and up the Rio Brasinho for some 
distance through the forest. Among the plants found hereabout may be named 
Mouriria guianensis, Aubl., Mapourea Martiana, Muell. Arg., Psychotria mhcrocea, 
Muell. Arg. ; Ancemopcegma sylvestre, a new^ Bignoniad with white flowers ; a new JDichcea 
{D.co?^nufa),ne3.v D. graminoides, Lindl.,from Guiana; Epidendrnm rariegatum, Hook., 
very common on trees overhanging the river ; another orchid, apparently an Epidendriim 
belonging to the section Amphiglottium, with the curious habit of producing new shoots 
from its persistent flower-scapes ; and Bandia Buiziana, DC, var. longlflora, again. Our 
second journey was up the Paraguay, with the object of reaching Diamantino if possible. 
The river runs through a belt of forest, beyond which, out of the reach of the precipitated 
moisture derived from fluviatile mists, there is much open campo. In the sliade of 
these riverside strips of forest grow plants often conspecific with or nearly related to 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 



!/.r 



those found in the great forest ; and the flora as a whole lias a large Amazonian element, 
not unalloyed, however, with a considerahle i)roportion of East-Brazilian types. It 
must suffiee to mention here Ingas [Inga uohilis, Willd., and I.nhdis, Mart.), tliellowers 
of which are often haunted by clouds of white and yellow butterllies belonging to the 
genus Callidryas ; a Sloanea (probably S. Mod'tmotoiczii, K. Schuni.) ; the beautiful shrub 
Miconia stenostachya, DC, with masses of white bloom variegated by the yellow of its 
anthers; the red -blossomed Passion-flower, Possijlorn vififolia. It. J5. K. ; Faramea 
bracteata, l^Qi\i\\., ?i Goyaz plant; Psychotrias; Ipomoea setifcro, Voir. ; the Amazonian 
Ci'oton Cajucara, Bentli. ; a VaniJJa with ensiform leaves, probably V. eusifolia^ lioUe — a 
New Granada species; the new Bronielia sj/lvicola; and Epidendrum imatophyllum^ 
Lindl., a Guiana orchid. 

We did not succeed in reaching Diamantino, and, on our return to Santa Cruz, were 
still without satisfactory intelligence concerning the prospects of the main ex])edition. 
It was now the rainy season, and the district was frequently visited by thunderstorms, 
which almost always came, so far as one was able to note, from the north, north-east, or 
north-west. In consequence of the rain, a number of plants growim;- in Wm (hici- 
situations suddenly burst into flower, so that collecting became very profitjible. Things 
continued in this state till Dec. 8th, when we received directions to come down to 
Corumba. On the way I made use of several small o[)portunities to collect plants while 
wood was being cut. AYe stayed a few hours at Villa Maria, now better known by its 
modern name, San Luiz de Caceres, and arrived at Corumba soon after midnight on 
Dec. 17th. 

Corumba is situated on the western — here, in consequence of a turn in the river, 
southern — side of the Paraguay, almost the 19th j^arallel of south latitude. Tlie town 
stands upon a bluff overlooking a wide expanse of low-lying country to the north, much 
of which is siiid to be under water during some period of the year. The geological 
structure is a hard limestone, the age of which Dr. Evans was unable to ascertain in 
consequence of the absence of fossils, but which is almost certainly Palaeozoic. Consider- 
able quantities of rain having fallen since the time of our passage up the river, the 
aspect of things vegetable was now very different, and many plants were preparing to 
blossom or were actually in flower. Several rough tracks run out from the city througli 
dense boscage, consisting for the most part of tall shrubs or small trees, larger trees being 
now and again met with. Fringing the river-side is a low-lying sandy area plentifully 
strewn with bushes and variegated with ponds and small pools, the haunt, I was told, of 
the Victoria regla, a statement I was unable to confirm. In one place this low-lying- 
land is about 400 yards wide ; elsewhere it is narrower, or the bluff may come close up 
to the water. At the time of our former visit, the greater part of this ground was imder 
water although it was the dry season — the fact being that during the up-country wet 
season the water is low at Corumba, and vice versa ; yet the seasons are synchronous, or 
nearly so, in the two regions — a striking testimony, this, to the extent of country drained 
by the Paraguay river. Walking by the river-side you will be surprised at the large 
amount of Croton scrub ( C nicifer, corumbeuslSy doctorls, all new species) exhaling its- 
peculiar aromatic odour, evidently very attractive to insects ; Cassias ( Cassia alata, Linn., 

2o2 



276 MK. SPENCER LE M. MOOKE— PHANEROGAMIC BOTANY 

and Tora, Linn.) are in full flower in January. DiantJiera paludosa, Sm., Ciiscuta ohtusi- 
Jlot^a, H. B. K., Tpomcea fistulosa, Mart., sometimes forming dense thickets, Leonotis 
nepetcpfoUa, R. Br., frequently occur, and Mom.ordiea Charanfut, Linn., spreads a literal 
carpet over large spaces of ground. Climbing the shrubs are the Amazonian Cenh^osema 
vexillatum, Benth., also Phaseohis lasiocarpus, Mart., Bignoniads, IpomcEas, and Passion- 
flowers ; and the Goyaba tree, here often attacked by the Loranthad Psittacmithiis cordatus, 
Blume, and covered with its red blossoms at this time of year, is a feature in the landscape. 
On the slopes facing the river and on the top of the bluff many plants will be found : such 
as Coccolobas ; Jatropha vitifolkf, Mill., in wearying iteration ; an orange-flowered Cereus ; 
Pereskia Bleo, DC, with its pale purple flowers ; Thyallis Laburnum^ a Malpighiaceous 
shrub new to science, reminding one, when blossoming, of our familiar Laburnum tree ; 
a fine Leguminous tree, Pterocarpus Pohrii, Vahl, with species of Seshania, Pauhinia, 
Psidimn, Panwoljia, Thevetia, Jiilocroton, and many others ; while of climbers we have 
Bignoniads, Teramnus rohibil'is, Sw. (like Centroscma vexillatum, Benth., an Amazonian 
species), a flow^erless Dioscorea with bright purple leaves, Phabdadenia Pohlii, Muell. 
Arg., Thlnouia sepium, S. Moore, &c. The pretty little Talinum crasslf'olium^ Willd., 
grows in masses upon the open ground, accompanied by a Cyphiacanthus, flowering 
specimens of which I, by some oversight, failed to secure. The streets of the town are 
enlivened by the brilliant blossoms of Ccesalpinia pulcherrima, Sw., planted here of 
course; and at the custom-house are some fine fig-trees {Ficus Elliotiana, S. Moore). 
Cratceva Tapia, Linn,, too, with its spreading leafy branches, affords most grateful shade. 
The river has its usual fringe of " camalote " mingled with grasses, Jussia?as, &c., and large 
islands composed of these plants float out from the neighbouring " riacho " or backwater 
called the Bahia de Caceres and become stranded along the shore, where the camalote 
is greedily eaten by cattle and i:)igs. 

Our stay at Corumba lasted six wrecks; it was varied by a short trip to Carandajinho, 
a small wood-cutting and cattle-raising station a few leagues up river. Here I found, 
upon marshy ground liable to inundation, large pink trusses of Phabdadenia Pohlii, 
Muell. Arg. ; Cassia aculeata, Benth., ^schynomene sensitiva, Sw., Byttneria campestris, 
S. Moore, and a Corynostylis, an Amazonian genus of which I saw no signs further 
up the river, were also gathered. In the ponds here were found the handsome aquatics, 
Thalia geniciilata, Linn., and Echinodorus panicitlatus, Micheli. Eestooning the trees on 
the river-bank was Macfadyenia laurifolia, Miers ; but I was able to do but little work 
here, and that only in the immediate neighbourhood of the boat, to which latter I was 
confined by indisposition during most of the time our visit lasted. 

We left Corumba in the Explorer on Feb. 1st, having a few days previously been 
joined by Lieut. Storm, who in the meanwhile had gone to Buenos Ayres with the view 
of placing himself in telegraphic communication with Europe, and had returned with 
intelligence of so unsatisfactory a character that no option was left him but to abandon 
the expedition. On our way down w^e called at Coimbra, picturesquely placed on the 
western bank of the river. Here we paid a visit to some limeston«; caves, which 
Weddell* tells us he also explored. To reach the caves you have to ninke your way 

* Ann. d. Sc. Xat. 3"" serie, xiii, p. 61 (in note). 



OF THE MATTO GKOSSO I-XPEDITION, 1891-92. 



■211 



through tall grass, as high as a man. This grass is Andropogon condenaaliis, H. 13. K., 
var. paniculatus. Growing in the grass you Avil] see here and tliere the J 11 amandft -like 
Thevetia Ucornuta, Muell. Arg., a Paraguayan species found as far to the north as 
Corumba, and on the river-bank you will iind Angelonta Gardneri, Hook., Teramiina 
volubilis, Sw., the curious Baillonia cmuibilis, Bocq., Tephrosla brevipes, Benth., Mimma 
cinerea, Veil., and grasses {Panicum^ Setaria, Imperata brasiliensis, Trin., and the tk ^> 
genus PogocJiloa), while the inevitable camalote swarms at the waterside. 

There is little more to tell. Soon after leaving Coimbra we reached the region of tlie 
Caranda forests ; dead logs of these trees were taken on board, for they yield excellent 
fuel. On Eeb. 3rd we were at Puerto Pacheco, a small settlement founded by tlu5 
Bolivians, but then in possession of a Paraguayan force. A short ramble in the H<>igli- 
bourhood yielded several plants of interest, including P/'osop/*" i-uscifoUa^ Griseb., a Chaci) 
type with very large and strong spines ; the Argentine Tabebuia nodosa, Griseb., a curious 
plant having Cuban aflB.nities ; Coccoloba paragnai'leusls, Linda u; Faapalum i^iinplcr, 
Morong; and new species of Wlssadiila, Mimosa, Morrenia^ Croton, and Julocrolon. 
But we could make no stay here ; our funds were exhausted, and it was necessary to 
hasten on to Asuncion; at that place, which we reached on Peb. 7th, the expcMlition 
was disbanded. 

Short Sketch of the Climatology of Matto Giiosso. 

In order that we may understand the origin imd nature of, and satisfactorily study 
the relations borne by the flora of a country to that of neighbouring regions, some 
knowledge of its climate, if it be not essential, is, at least, manifestly advantageous. I 
trust therefore that no apology is needed for presenting to the reader the following short 
sketch, to a large extent compiled from various available sources. Nor have I refrained 
from adding a few observations made by myself during our expedition. It must be 
understood that at Santa Cruz, the climatology of which is quite unknown, with the 
exception of a couple of ordinary thermometers, we had no recording instruments 
whatever ; that we went there in the expectation of making but a short stay ; and that, 
for all we knew, the order for our departure might arrive any day. Hence it did not 
seem worth while to commence a series of observations which must necessarily have 
proved most imperfect, and were liable to cease at a short notice. 

The dry season lasts from March or April to September, and during this period storms 
and rain are exceedingly rare. They do sometimes occur, however ; thus, on the night 
of Aug. 11th, when we were nearing Cuyaba, on our upward journey, a violent thunder- 
storm, accompanied by a perfect deluge of rain, broke over the country. At the beginning 
of September a premonitory storm or two pass over, soon followed by the regular wet 
season. Por a tropical country the rainfall is not high. Dr. Morsback *, as the result 
of three years' observations, got an annual average of 1166 mm., that is between l.j and 
16 inches, only a trifle greater than the rainfall at Rio. The reason for this must, I 
think, be sought in the geographical position of the city, and in the nature and direction 
of the prevailing winds, which blow mostly from the north, north-west, and north-east. 

* ' Revista do Obaeryatorio,' Rio, 1890, p. 91. 



278 ME. SPENCER LE M. MOORE — PHANEROGAMIC BOTANY 

These winds bring moisture from tlie low-lying Amazonian region, moisture which is^ 
however, largely intercepted by the high plateaux lying to the north of the province. 
On the other hand, south winds, generated in or having to pass over the dry Pampas, 
will not carry much moisture with them. To judge from our experience at Santa Cruz 
and at Corumba, there is much fine weather during the wet season, several days of 
brilliant sunshine sometimes intervening between the storms ; and this is borne out by 
the observations of Lt.-Col. Americo llodrigiies de Vasconcellos * at Cuyaba, this 
observer recording for January and Eebruary 1889, the middle of the wet season, only 
11 and 15 rainy days respectively. 

The main feature in the temperature is its liability to great and sudden variations. I 
may cite in illustration the testimony of Dr. Joao Severiano da Fonseca f, who accom- 
panied as doctor the Brazilio-Bolivian Boundary Commission of 1875-78. On June IDth 
of one of these years, on the Cuyaba river, he noted a temperature of 35 C. at 
2 o'clock P.M., and by 10 o'clock that night the mercury had fallen to 7°'5 C. : that is, an 
oscillation of more than 50 degrees E. within eight hours ! This is, of course, an 
extreme case. On another occasion he recorded a sudden fall, during a hailstorm, of as 
much as 18°* 7 C. In the highlands frosts are reported to occur sometimes, nearly every 
year, between June and September, and the Indians are occasionally frozen to death ; 
moreover, the young coffee-bushes arc then liable to destruction. Fonseca also says that 
one morning, near the 16th parallel of south latitude, he saw the ground covered with 
hoar-frost and the puddles frozen over, and this at the level of the river ! During 
Dr. Morsback's three years at Cuyaba, the temperature varied between 4!l°"9 C. and 
7°*2 C. This temperature of 11°'9 C. (upAvards of 107° T.) is, so far as I am aware, the 
highest on record for the province. Our highest at Santa Cruz was 38°"5 C. on Monday, 
Oct. 12th. 

As an instance of the daily range of temperature, I take the opportunity of citing 
the following figures, deduced from De Vasconcellos's observations at Cuyaba for 
July 1888 t :— 

7 A.M. 10 A.M. 1 P.M. 4 r.M- 

Average monthly temperature in the shade 19°-06C. 25°'98C. 30°-63C. 80°-81 C. 

The hottest period of the day is usually some time between 1 and 1 p.m. At Santa 
Cruz the maximum was generally reached at about 3 p.m., though exceptions often occur, 
the mercury at 6 o'clock sometimes standing higher than at any earlier hour of the day. 
Moreover, the morning hours are occasionally the hottest. 

The oscillations of the barometer are but slight ; the maximum monthly oscillation, 
as set forth in the tables to which I have had access, amounts to rather more than 
13 millimetres (Cuyaba, August 1888, De Vasconcellos), but this is quite exceptional, as 
the subjoined figures will show : — 

* ' Revista do ObsciTatorio/ Rio, 1880, pp. OG and 8-'$. 

t • Viagem ao redor do Brazil,' Bio, 1881. 

t ' Kevistu do Observatorio," liio, 1888, p- 185, 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 



279 



Ohsernations at Cuyabd made hij Lt.-Col. Amerlco Hodriffiies de Fasconcellos 



>* 



Month. 
July 1888. 



„ , , /-Max, 751-73 mm. 

barometer reduced \ ,,. 

toO»C.: '■-•r'-?" 

V. Mean 748-34 „ 

m . • .1- rMax.36°-3C. (lowest day max. 3 1°-9C.). 

Temperature m the \ ,,. .^ - ^ ,, - , ' , . ^ ^. 

^ , J ^Mm. 14°-5C. highest night mm. 20°-9C.). - 

'^^^''- (Mean 28-0 C. 

Average tension of") 

^ [ 12-76. 

aqueous vapour : J 

Relative humidity : 539. 

Wind: variable on 17 days; N. or N.VV. 8 days; S. or S.E. 6 days. 

No rain fell during the month. 



August 
1888. 



„ ^ - , f Max. 751-44 mm. 

liaroracter reduced \ ,^. 

^ Mean 746-24 „ 

_, . - ., /-Max. 39°-9C. (lowest day max. 17'*-7C.). 

Temperature in the y ,,. „ •' „ ' 

^ Mm. 10°-5C. (highest night min. 2i°-6C.). 

C Mean 30^-2 C. 

Average tension of-\ 

It Q.QQ 

aqueous vapour : J ' ^ ' 
Relative humidity : 44'6. 
Wind : "N., N.E.^ or N.W. during greater part of month; S. or S.W, 

on 4 days. 
Rainfall : 0*4 mm. on one day only. 



September 
1888. 



_, , , f Max. 74748 mm. 

liarometer reduced \ ,_. „,^ ,^ 

V Mean 747-48 „ 

^ . , /-Max. 40°-9C. (lowest day max. 26°-9C.). 

Temperature in the \ ,,. „ ^. ^ . , . , . « \^ 

*^ < Mm. 20°-5 C. (highest night mm. 27°- 1 C). 
SiiaQS c / . 



Mean30°-6C. 



Average tension of 



1 20-^ 



aqueous vapour 
Relative humidity : 63-8. 

Wind : N., N.E., or N.W. 14 days; S.W. 2 days; variable 14 days. 
Rainfall : 52*07 mm. Rainy days 10. 



Kevista do Observatorio,' 188S and 1S89, jpassim. 



280 



ME.- SPENCEE LE M. MOOEE— PHANEEOQAMIC BOTANT 



Month . 

October 
1888. 



^ f Max. 746-96 mm. 
Barometer reduced \ .,. -,„^,„ 

toO°C. : /^^ „^^^„ " 

iMean 743'27 „ 

. , /-Max. 37°-4C. (lowest day max. 28°-9C.). 
Temperature m the ^ ^^.^ ^^..^ ^ ^^.^^^^^^ ^,^^^ ^^ ^^,^^ ^^ 



shade : 
Average tension of 






Mean 31°-0 C. 



:*■} 



21-60. 



aqueous vapour 
Relative humidity : 66. 

Wind : N.,N.E., orN.W. 1."^ days; S. or S.E.2days; variable 14 days. 
Rainfall : 38'8 mm. llainv davs 10. 



January 
1889. 



-r, ^ J , /-Max. 747-10 mm. 
Barometer reduced \ ,,. 

toO°C.. ]«'"• ^^-O" " 

^Mean 744' 30 ,, y 

„ ^ . ^, r Max. 39°-9C. (lowest day max. 29°-4C.). 

Temperature m the \ , ,. „ 

^ ^ Mm. 22°-l C. (highest night min. 25°-4 C). 



shade : 



^ Mean 30°-4 C. 
1 21-87. 



Average tension of 

aqueous vapour 
.Relative humidity: 68*7. 
N.W. wind predominant. 
Rainfall: 145*7 mm. Rainy days 11. 



February 
1889. 



March 

1889. 



, rMax. 747- 19 mm. 
Barometer reduced \ ,^. 

toO»C.: I^- Ifj: " 

vMean 744-67 ,, 

^ ^ • , r Max. 37°-9C. (lowest day max. 29°-9C.). 

lemperaturc m the V ^. ^ ,n ^ ^ A . , , 

. . j Mm. 22°-2C. (highest night rain. 22°-2C. 

^ ■ (-Mean30°-7C. 

Average tension of) 

J'21-90. 
aqueous vapour : I 

Relative humidity : 68' 0. 

Predominant wind : N.W. 

Rainfall : 1388 mm. Rainy days 15. 



T, , 1 1 r Max. 7479 mm. 

Barometer reduced \ ^ ,. 

toO°C • l^^'"'- ^40-99 „ 

t-Mean 744-62 „ 

Temperature in the C^'^' ^^^''^' ^^°^^^* ^^^ «^^^- ^^'^C')- 
shade ■ / "^^^'^' ^^°'^ ^' (^^g^^^* ^^S^* ™in- 24°-6 C. 

^Mean 29°-8 C. 

Average tension ofl 

>■ 21-44 
aqueous vapour : J 

Relative humidity : 69-8. 

Predominant wind : N.N.W, 

Rainfall not given. 



OF THE MATTO GEOSSO EXPEDITIOX, 1891-{)2. 



281 



Month. 
April 1889. 



^Max. 748-58 mm. 
Barometer reduced \ ,,. 

toO'C: if"- "■*'-^- " 

"-Mean 715-50 „ 

^ X . , r-^la^"- 37°-8C. (lowest day max. 21° C). 

Temperature m the \ „ ' 

sh d • ; '^^"^' ^•"* ■^^- ('"o''tJst night min. 23°-5C.), 

^Mean29°-5C. 

sion of"| 

MO- 95. 
aqueous vapour : J 

Relative humidity : 66'8. 

Predominant wind : N. 

Rainfall not entered. 



Average tension of~ 



It will here be useful to append Pr. Morsback's * results, obtaiaed at the cost of 
three years' observation at Ciiyaba, from July 1881 to June 1887 : — 

/-Highest mean 750-757 mm. 
^ Lowest mean 738-710 ,, 
V. Annual mean 741-745 „ 

^m": r-S a Monthly „ean(;"«''-' f/'OC. Sept.-May. 

Mean temperature at 7 a.m. 21°-9C. 
„2p.m. 33°-77C. 
,, 9 P.M. 23°06C. 
Average annual rainfall : 116G mm. Mean of rainy days : 85. 

A few figures ha^e recently been obtained for Corumbd ; they are here recorded : — 

Observations made at Corumbd hy the Matto Grosso Illlitary Eagineering Commission 

voider Capt. Jorge dos Santos Almeida i. 



Barometer 



Temperature in the 
shade : 



Month. 
November 
1889. 



/-Max. 751"0 mm. 
Barometer: j Min. 7409 ,, 
^Mean 746*4 ,, 
Temperature in the rMax. 36°-4 C. (lowest day max. 26°-5C.). 

shade : \Min. 17°-5 C. (highest night min. 27°-4 C). 

Rainfall : 1899 mm. Rainy days : 13. 
Predominant winds : N.E. and N.W. 



December 
1889. 



Barometer : 



rMax. 75 ri ram. 

-j Min. 739-3 „ 

^Mean 745-9 ,, 
Temperature in the f Max. 35°-8 C. (lowest day max. 23°'4 C). 

shade : iMin. 18°-3 C. (highest night min. 27°-5 C). 

Rainfall : 240-9 mm. Rainy days : 15. 
Wind : Generallv N.E. or N.W. 



* ' Kevista do Observatorio,' Rio, 1 890, p. 91. 
t Ih'uL 1890, pp. 30, 47, 63. 
SECOND SERIES. BOTANY, VOL. IV. 



2p 



282 MR, SPENCER LE ^L MOORE— PHANEROGAMIC BOTANY 



Month, 

January 

1890* 



/-Max. 753-3 mm. 
Barometer : < Min. 742-0 „ 
(-Mean 747-5 „ 
Temperature in the /Max. 33°-8 C. (lowest day max. 24°-7C.). 

shade : iMin. 19°-8 C. (highest night min. 27°-0 C). 

Rainfall : 150-5 mm. Rainy days : 11. 
Wind : Predominantly N.E. or N.W. 



Here follow a few figures obtained by myself at Santa Cruz during the few days of 
my confinement to the boat from indisposition, which prevented collecting : — 

Shade Temperatures at Santa Cruz, Nov. 1891 (Centigrade scale). 

Nov. 19th. P.M. 1, 33°-5 ; 2.30, 34°; 3.0, 35°-5 ; 4, 35°-3; 4.30, 35°-3; 5, 33°-5 ; 5.30, 29°-7. 
f A.M. 7, 26°; 8, 26°-8 ; 9.15, 29°; 10.15, 30°- 5 ; 11.15, 32°-3. 
^"Ip.m. 12.15, 33°; 1.30,34°; 2.30, 35°-5 ; 4, 35°; 5, 34°-8; 6, 32°-5 ; 7, 29°. 
/-A.M. 6, 19°; 7, 19°; 8, 26°-8; 9, 30°-4 ; 10.30, 34°'2 ; 11.30, 35°-8. 
Nov. 21st. ^ P.M. 12.30, 35°-2; 1.30, 35°; 2.30, 36°-5 ; 3.30, 36° 2; 4.30, 32°; 5.30, 28°-8 ; 
^ 6.30, 27°-3. 

A.M. 8.30, 23°-5; 9.30, 24°-7; 10.30, 28°-8; 11.30, 29°-2. 



N '>7th i'^'^^' ' ^^ ^^' *^-^^^ 24"-7; 10.30, 28"-8; 11.30, 29"-2. 

"Ip.m. 12.30, 31°-2; 2, 32°; 3, 32°; 4, 30° 5, 26°-4; 7.15, 25°. 

A.M. 6.45, 20°'6 ; 7.45, 21°-3; 9.15, 24°-5; 10.15, 27°-8; 11.15, 30°-4. 

31°-6; 2,32°; 3.15,32°-5; 4.15, 32°-2; 5.30, 31°-8 ; 6.45, 30°-4; 8, 26°- 3. 
A.M. 6.30, 20°'2; 7.30, 22°- 6 ; 9, 26°-8; 10-30, 31°-2; 11.30, 33°-5. 



Nov. 28th. r-'^-^-^ 
Cp.m. 1, 



N ' 29th (^'^^■^•'^^^ '^^ '^'y 7.30, 22"-6; 9, 26"-8; 10-30, 31"-2; 11.30, 33"'5. 

■ 1p.m. 2.15, 33°-5; 3.15, 34°; 5.15, 34°-5; 6.15, 34°; 7.15, 30°- 2 ; 8, 28°-4. 

^"i P.M. 12.30, 35°; 1.30, 35°; 2.30, 33°-8 ; 3.30, 34°' 8 ; 5, 35°-7; 6, 35°; 7, 33°-3. 

Weather at Santa Cruz. 

Nov. 16th. Fine day with fresh breeze ; no rain ; distant thunder. Temperature at 1 o'clock 32°*2C. 
1 7th. Fine day, fresh breeze. A little rain in morning. Temperature at 2 o'clock 34° C. 
18th. Distant thunder early afternoon : thunderstorm from N.E. early evening. Temperature 

shortly after 2 o'clock 36° C. 
19th. Distant thunder heard in the south at 4 p.m. 
20th. Overcast morning; fine afternoon; no rain; no thunder. 
21st. Fine morning; short rainstorm in afternoon, with a little distant thunder. Lightning 

in S.E. during evening. 
22nd. Heavy thunderstorm from N.E. lasted several hours this morning, 
23rd. Rain during night, with some lightning. 

24th. Heavy rain for upwards of an hour in morning. Fine afterw ards. 
25th. Heavy shower from S.W. during morning. No thunder. 
26th. Fine day ; no rain ; no thunder. 

27th. Thunderstorm from S.E. between 7 and 8.30 a.m. with heavy rain. Lightning in 
evening. 



By a clerical error this is given as a result for Dec. 1889. 



OF THE MATTO GKOSSO EXPEDITION, lSi)l-92. 283- 

Nov. 28tli. Brilliant da}- ; no rain ; no thunder. 

29th. P'ine day, only very slight shower of rain. 

30tli. Early part of evening stiflingly hot; thunderstorm burst soon after 8 p.m. 
Dec. 1st. Fine day. Highest temperature .'J7''-;3 C, just before 5 o'clock. 

2nd. Fine day, without rain. Highest temperature precisely as on prcccfling day. 

3rd. Fine morning. Afternoon a distant thuuderstorni, with little rain. 

The above figures sliow the diuvnal ranj-i^ of 1 einperature, and briiiu^ out clearly the 
fact that, at least in November, the air is hottest after 2 p.m. ; indeed, on Nov. 29th and 
30th the maximum was not reached before 5 o'clock, and, remarkablv cnouirh, at 
7 in the evening of the hitter day the mercury actually stood higlier than at any hour 
of the two days preceding the 29th. A cursory inspection of the weather column 
will show a confirmation of the statement already made, that int(U'vals of brilliant 
weather may be experienced during the wet season ; thus, for the week ending Dec. 2nd 
we find, but for a couple of thunderstorms, fine weather prevailing throughout. 

The Botanical (lEOGjtAriiv oi- Hkazil. 

For the earliest generalizations of any practical value bearing upon tlic l)utauical 
geography of Brazil we must go to Von Martins, whose views were adopted in those 
earlier volumes of his great ' Flora Brasiliensis ' which were published before tlie 
eminent traveller's death, and are still recognized in the various parts issued and now- 
issuing. Martins distinguished the low watered plains of Brazil as the Naiad r(;gion, 
while to the tropical mountain forests of llio de Janeiro and the eastern coast he 
applied the term Dryad region. The back parts of the provinces of Bahia and 
Pernambuco have an exceedingly hot and dry climate, which the uj^lands of Ceara, 
Piauhy, and the neighbouring provinces also share ; this is the Hamadryad region. 
An Oread region is also distinguished; it comprises the uplands of Minas Geraes, 
Goyaz, Matto Grosso, and Sao Paulo. The small portion of the country lying south 
of the Tropic of Capricorn Von Martins distinguishes as the Napcca region. In the year 
1872 appeared Grisebach's 'Vegetation der Erde,' a work which may fitly be called 
the first modern attempt to grapple with the problems pertaining to the distribution 
of plants over the whole globe. For the bulk of Brazil Grisebach recognizes only 
two regions ; his extra-tropical part he includes in his Pampas region. Tiie northern 
of these two regions, or Amazon region, corresponds with the Naiad region of Von 
Martius ; its northern limit runs nearly j)arallel with and close to the Equator, while 
in the south-east its boundary includes Cearii, Piauhy, and Maranhao, and is continued 
westward to the neighbourhood of the Andes, but without anywhere reaching the 
limit of 10 S. latitude. The rest of tropical Brazil Grisebach unites into one large 
Brazilian region. Engler's * views are somewhat different. He objects to the inclusion 
in the Amazon region of the dry Piauhy, Ceara, and Maranhao uplands, and he sees 
no reason why the whole of the country drained by the Amazon and Tocantins rivers 
and their feeders should not be included in one province t, with which he throws 

* ' Versuch einer EntwickelungsgescMchte der Pflanzcnwelt." 

t He excludes, of course, the strip of country bordering on the Andes ; this constitutes his Subandine province. 

2p2 



284 ME. SPENCER LE M. MOOEE— PHA^^EEOGAMIC BOTAXY 

in Guiana, a country united by Grisebach with Venezuela and parts of Colombia 
and Central America to form bis Cis-Equatorial region. In this way Engler defines 
a North'Bi^azil- Guiana province. Besides the rest of Brazil, his South- Brazilian 
province comprises Uruguay, Paraguay, the Argentine provinces of Entre Bios and 
Corrientes, Eastern Bolivia, and the comparatively high-lying country immediately 
westward of the Cliaco. This latter lie considers doubtful land, possibly with a 
predominating Pampas flora. Engler's North-Brazil- Guiana province — taking in, it 
may also be said, the Venezuelan savannas — thus includes almost the whole of Goyaz, 
as well as the little-known northern part of Matto Grosso ; the southern boundary of this 
province will be the watershed between the Xingu, Tapajos, and Guapore rivers to 
the north and the Paraguay and its feeders to the south. Erom Drude's * Amazonas 
resrion the hiijhlands of Guiana are excluded, but all the ^.E. of Brazil, including the 
littoral almost to the 20tli parallel, forms part of the region whose southern limit 
practically coincides with that of Engler's North-Brazil-Guiana province. All the 
south of Brazil, with Uruguay, Argentina east of the Parana, the country behind 
the Chaco, and Eastern Bolivia together make up the Parana region of Drude, which 
thus largely corresponds with Engler's South-Brazilian province. 

There can, I tbink, be but little doubt that Engler's views are founded upon reason. 
Bounded on the north-east by tlie uplands of Maranhao and Piauhy, which, to say 
nothing of the tract of country lying still further to the east, are scarcely to be 
considered as having much in common with the moist Amazon forest-zone — in the 
east and south-east separated from Minas Geraes and Southern Goyaz by several 
ranges of high hills, such as the Serras do Duro, da Tabatinga, and divisoes do Bio 
Claro,^ — orograpbical data are certainly on the side of the Berlin Professor. The weak 
point in his position consists in the fact that the watershed separating the Paraguay 
head-waters from those of the Xingu, Tapajos, and Guapore is a remarkably low one ; 
well, therefore, might the thoughtful student doubt whether that watershed really 
marks the boundary between the two great provinces, and, as will shortly be shown, 
the collections which I have just brought back prove such a doubt to be warranted. 
A few words will first be devoted to the flora of Cuyaba and the Chapada Plateau, 
after which we will consider the flora of Santa Cruz and, more interesting still, that of 
the forest-region situated to the north of that settlement. 

Elora of Ctjyaba and the Chapada Plateau. 
The city of Cuyaba lying fifteen and a half degrees south of the line, cursory inspec- 
tion will convince the reader that it is well within the limits of Engler's South-Brazilian 
province as of Drude's Parana region. "We should therefore expect its flora — especially 
the plateau portion thereof — to show much similarity to that of some of the eastern 
provinces, especially the upland country of Minas Geraes and Southern Goyaz. At the 
same time the provincial boundary is not very far off, the distance between Cuyaba and the 
sources of the Arinos being only about a hundred miles. Consequently no surprise should 

* ' Die Florenreiehe der Erde.' Eeprinted from Petermann's ' Mittheilungen,' 1883-84. 



OF THE MATTO GROSSO EXPEDITION, 1801-92. 285 

be felt at stray plants characteristic of the Amazonian country turning- up near Cuyaha ; I 
did not meet with many such plants, but there is a certain ])ercentiige of that element in 
the flora, as will directly be shown. 

For the purpose of this and the following chapters, as also in the systematic portion 
of the present memoir, I have employed certain geograj)hical expressions which must 
now be explained. The term " Bras. Or. " is used to denote the whole of the eastern and 
south-eastern part of the country, from Ceara and Piauhy to llio Grande do Sul, including 
Bahia and the neighbouring littoral provinces, Minas Geraes and the soutlnn'n end of 
Goyaz. By " Amazonia " I mean the whole of the Amazon and Tocantins basin witJi 
the exception of the former's extreme \^estern end, which comes within the Subandine 
province; excluding, too, Northern and Central Goyaz, a district in my opinion certainly 
forming part of the North-Brazil-Guiana province, but kept distinct as lending to sliow, 
more satisfactorily than would otherwise be the case, the relations of I he Alaito Grosso 
floras to the flora of different portions of the above-named j)rovince. For the same 
reason Guiana — restricted to English, French, and Dutch — receives separate mention. 
The other districts are ])olitical, as it is sometimes extremely difllcult if not impossil)le to 
discover of what botanical province a given plant is a denizen — a remark of special 
reference to the north-western 2)arts of the continent. 

With a Aiew to making the following statements as wide in their applicjition as possible, 
T liaA'C diligently sought out the affinities of the species considered to be new, entering 
the result within brackets at the end of each description, but in these cases using italics. 
On the other hand, the known distribution of a species alreiidy described is given in a 
similar way, but without italics. This method of ascertaining the affinity of a new species, 
and considering such affinity when found as evidence of phytogeographical value, has 
much to be said in its favour, especially in the case of one who has exhaustively mono- 
graphed the genus to which is referable any given plant so dealt with ; though, in less 
favourable circumstances, explicit reliance upon an author's conclusions can hardly be 
expected, however laboriously he may have reached them. The only alternative is, when 
tabulating one's results, to omit all the new species from the calculation, an unsatisfactory 
method when, as in the present case, a large number of species are new. 

Of the flora now to be considered, as of most of the other floms, the chief factor is the 
Tropical American, consisting, that is to say, of species having a wide range through the 
American, or at least South- American tropics. No less than thirty-five per cent, of the 
flora is of this nature *. There is also a large contingent (twenty-six per cent.) of types 
occurring both in Eastern Brazil and in some part of the North -Brazil-Guiana province. 
The element next in importance is the East Brazilian (twenty-four per cent.) ; while the 
North-Brazil-Guiana province is represented to the extent of only ten per cent. Other 
usmall elements of the flora are the East-Brazil-Paraguay, the East-Brazil-Goyaz-Paraguay, 
and the East-Brazil- Amazonia-Paraguay, together making up between two and three 
per cent. In accordance with Engler's scheme, the first of these should be included in 
ihe South-Brazflian flora, and the two latter dealt with in the enumeration of species 

* Here and elsew here decimals are omitted. 



286 ME. SPENCEE LE M. MOOEE — PHANEEOGAMIC BOTANY 

common to the two provinces into \Yhicli the Berlin Professor divides Brazil. Making- 
the reqnisite modifications in the foregoing statement, the flora of Cuyaba and the 
Chapada Plateau may he said to be made up of the following factors : — 

Diffused Tropical American 35 per cent. 

Common to tlie two Brazilian provinces 28 „ „ 

South Brazilian 25 „ „ 

Xorth-Brazil-Gniana 10 „ „ 

Of this last factor it may be added that — 

r Types restricted to Amazonia constitute 40 per cent. 

-^ Types restricted to Goyaz „ 40 „ ,, 

LTypes of wider distribution ,, 20 „ „ 

We have already seen the rainfall of Cuyaba to be, for a tropical country, but scanty ; 
moreover, the dry season, while it is of long duration, is very seldom interrupted by 
storms. The vegetation of the open country, as distinguished from that of the moister 
valleys, should therefore show some adaptation to xerophilous conditions. The swollen 
rootstocks, characteristic of many Matto Grosso herbs and small shrubs, are undoubtedly 
in point here ; an examination of the underground parts of trees and the larger shru.bs 
would probably show this adaptation to be very common. Coriaceous leaves also are the 
rule ; and the trunks of trees are often thickly covered with cork, apparently functioning 
as a hindrance to transpiration. The same end is often served by great diminution 
in the number of leaves, so that it is by no means unusual to see shrubs perfectly leafless 
except immediately beneath the inflorescence. The assiimption by leaves of a position of 
least insolation is rarely observed : Xylopia grandifiora, A. St.-Hil., which has them verti- 
cally pendent, and Myrcia amhigim^ DC, are the best cases noted by me. Spiny plants 
do not occur in any great quantity. 

Among noteworthy North-Brazil-Guiana species, or species having their nearest affinity 
with such, may be cited Banisteria constricta, Griseb., BauJdnia cumanensis, H. B. K., 
Inga nobilis, Willd., and the new species Guatteria sylcicola, GUanthus collinuSy Laden- 
bergia chapadensis, and Ajouea pruinosa. Among genera abundantly represented in 
Eastern Brazil, but here dying down, none is more noteworthy than the Labiate Syptis. 
Of this genus no less than a hundred and sixteen species are peculiar to the South- 
Brazilian province, while those endemic in North-Brazil-Guiana number but five, and of 
the forty-five species common to both provinces only six reach Amazonia. Similar facts 
may be cited for the genus Begonia and others ; they might well lead us to suspect, as is 
actually the case, that at Cuyaba we are near the limits of a botanical province. 

Flora of Jangada. 

The flora of this place may be taken as representing that of the large district to the 
north of Cuyaba lying between the Cuyaba and the Paraguay rivers. The various factors 
of the flora, as shown in my collection, may be stated thus : — 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 287 

Diffused Tropical American 27 per cent. 

Common to the two Brazilian provinces 37 ,, „ 

South Brazilian 21 „ „ 

North-Brazil-Guiana 8 „ „ 

Comparing this with the Cuyaba flora, one notes diminution in the Tropical Americjin 
as also in the Nortli-Brazil-Guiana element, compensated, however, by an increase in 
the number of types common to the two Brazilian provinces. The South-Brazilian 
contingent is approximately identical in the case of both iloras. One may conclude that 
the present flora, like that of Cuyaba, is markedly South Brazilian in character, and 
except for the high percentage of species common to the two r>razilian floras, yields but 
slight indications of the fact that at Jangada we are in the neighbourhood of another 
floristic province. 

Plora of Santa Cruz and the Neighbourhood. 

Por the purpose of the following computation, I have included under this hcml not 
only the plants found in the open country at and close to Santa Cruz, but also those 
gathered during our expeditions above that settlement up the rivers Paraguay, dos 
Bugres, and Brasinho. Since it is highly probable that the two latter flow through the 
primeval forest at least in the upper part of their course, the plants gathered upon their 
banks should perhaps have been added to the primeval forest list. It is, however, very 
difficult, in passing up a river with densely wooded banks, to say how far the thick 
boscage extends, and whether or no the country soon becomes more open. Mr. Holden, 
who, at our halting-places, used to push his way through the littoral " matto " in search of 
sport, is of opinion that, at least in the case of the Paraguay, open campo is soon reached. 
This, however, can scarcely be correct so far as the Brasinho is concerned ; for although 
in oiu' forest expedition we did not strike this river, yet several of its feeders, such as the 
San Pedro, Alegra, and Palmitar, were forded, and the banks of these streams showed 
striking similarity to those of the Brasinho. Moreover, the fact of ipecacuanha being 
found near the latter river points to the existence of much forest-land there. 

The flora is composed of the following items : — 

Diffused Tropical American 37 per cent. 

Common to the two Brazilian provinces 28 „ „ 

North-Brazil-Guiana li> „ „ 

South Brazilian 13 „ „ 

If the reader will take the trouble to refer to the Cuyaba figures, he will find the 
second item precisely the same in both cases, and the first approximately so. The per- 
centage of South-Brazilian types has, however, fallen from twenty-five to thirteen, while 
the North-Brazil-Guiana j)ercentage, at Cuyaba but ten, is now nineteen ; or, speaking 
roughly, one element has undergone lessening and the other increase to the extent of a 
hundred per cent. 

It may, perhaps, not be wholly uninteresting to show the distribution of the species 
common to the two provinces, as of those endemic in the North-Brazil-Guiana province. 



288 ME. SPENCEE LE M. MOOEE — PHANEEOGAMIC BOTANY 

Of the species common to tlie two Brazilian provinces we have : — 

From Govaz and Bras. Or. (or with nearest affinity to such species) 38 per cent. 

From Amazonia, Goyaz, and Bras. Or. „ „ 23 

From Amazonia and Bras. Or. „ „ 13 

From Goyaz, Bohvia, Peru, Colombia, and Bras. Or. „ „ 10 

From Goyaz, Guiana, and Bras. Or. „ „ 7 

From Bohvia and Bras. Or. „ „ 4 

From Guiana and Bras. Or. ,, „ 2 

From Goyaz, Amer. Cent., Bras. Or. „ „ 1 

From Goyaz, Paraguay, Bras. Or. „ „ 1 

Of the T^orth-Brazil-Gniana species there occur : — 

In Amazonia and G oyaz 76 per cent. 

In Guiana 13 „ „ 

In Colombia 5 „ ,, 

In Peru 3 „ „ 

It must not be forgotten that the inclusion in the Santa Cruz flora of plants found 
growing on the banks of the Kio dos Bugres and Brasinho tends somewhat to enlarge 
the North-Brazil-Guiana factor in that flora. In spite of this, one may consider the 
flora as containing a comparatively large North-Brazil-Guiana element and a smaller 
South Brazilian. At Santa Cruz, therefore, we have a mixed flora in which North- 
Brazilian plants somewhat predominate, and the assertion that Santa Cruz ought to be 
included within the l^ounds of the North-Brazil-Guiana province will, it is trusted, be 
deemed in no way unreasonable. 

The statement just made may be supported by some additional data. One of the 
most singular facts in the whole range of phytogeography is the South- American distri- 
bution of Cycads, an order which, although in the nortliern hemisphere reaching to 
Florida and the Bahamas, is nowhere represented in the eastern part of South America. 
This is the more remarkable because in the Old World we find Cycads as far north as the 
thirty-third parallel, while in Australia and at the Cape they almost or quite attain to 
the thirty-fifth degree of south latitude. Moreover, most of eastern South America is of 
high geological antiquity, so that a priori one would certainly expect to find Cycads 
there. But whatever be the reason of their absence, the occurrence of a Zamia [Z. Brong- 
niartii., Wedd.) at Santa Cruz close to the fifty-seventh parallel of longitude is worthy of 
notice. This species has been found in Eastern Bolivia at St. Xavier (lat. 16° 10'), a place 
about a degree further south than Santa Cruz, and this, so far as is known, is the southern 
limit of the order in America. The nearest neighbour of this Zamia is Z. Poeppigianay 
Mart. & Endl., found by Poeppig on the banks of the Tocache river in Eastern Peru, just 
within the bounds of the Subandine province ; but, as Eichler * observes, it will probably 
be discovered within the Amazonian region, a region within which no Cycad has yet been 
met with. But although Cycads are not known from the North-Brazil-Guiana province,, 
as defined by Engler, the occurrence of Zamia Poeppigiana near the boundary of that 

* Mart. Fl. Bras. iv. pars i. p. 416. 



OF THE iVIATTO GROSSO EXPEDITION, 1801-92. 289^ 

province and of other Cycads further to the north-east, coupkMl witli tlic entire absence 
of the order from Eastern South America, justifies us in view ini,' the existence of a Zmnia 
at Santa Cruz as an indication of a North- Brazil-Guiana afliiiilv as distinjjuislied from a 
South-Brazilian affinity in the Santa Cruz flora. 

Negative evidence is also of value in considerinj^ the nature of a flora ; and although 1 
do not propose to deal with this part of the subject at all exhaustively, a {aw facts may 
here be mentioned. I saw no species of Begonia at Santa Cruz or in the neiglil^ourliood *. 
Now, of eighty-three species of Bcfjonia enumerated by Alplionse de Candolle in tlii' 
fourth volume of Martius's ' Plora Brasiliensis,' all are East Brazilian except one Amazonian 
and one from Matto Grosso. Among Malvaceae the absence of the genus Ahutilon — a 
genus abundantly represented in the South-Brazilian province — is worthy of remark, as 
is also the paucity of Compositd', which in the same province appear in great force, both 
generically and specifically. Neither is the rarity of Labiatea^ witliout signiticance, of 
which order several genera and many species are endemic in Eastern Brazil : not one of 
these genera, such as Cunila, Hedeoma^ Keithia^ Glechon, was met with at Santa Cruz. 
Moreover my collection from Santa Cruz does not contain any representative of such 
preponderantly South-Brazilian genera of Scrophulariacea^ as Angelouia, SchtccHfcia, 
JBriwfclsia, or Stemodia. The Gesneras, too, are wanting at Santa Cruz; according to 
Hanstein f, Eastern Brazil claims nearly SO per cent, of Brazilian species, the localities of 
the remainder being doubtful. On the other hand the genus Drymonia, with head- 
quarters in Costa Hica, is endemic at Santa Cruz. 

The occurrence at Santa Cruz of the ne^v genus Brosimo})sis is an indication of the 
affinity of its flora with that of Amazonia, where several small or monotypic genera of 
Artocarpese flourish, BrosiniNm Gaffd/e/iand/'/, Tree, however, a characteristic East- 
Brazilian plant, is also met with at Santa Cruz. Unfortunately most of the orchids there 
were out of flower during our stay, but the short list of those in flower is very suggestive 
of the flora's northern affinity. Thus the Amazonian and Guiana Cattleya superbo, 
Schomb., flourishes at Santa Cruz, as also do Epidendrum imatophjlhnn, Lindl., a Guiana 
species, Notylia hisepala, S. Moore, of which N. Huegelii, Beichb. f ., from Mexico, seems 
the nearest aUy, and Dlch(ea coniuta, S. Moore, a species coming nearest to the Guiana 
I), graminoides, Lindl. A curious narrow-leaved Vanilla, which, when its flowers are 
known, will probably prove identical with V. ensifolia, Bolfe, from New Granada, should 
also be noticed. Lastly we have two remarkable grasses : one of these is the lowly 
Luziola pusilla, S. Moore, a species found also in Guiana ; the other, a new species of 
Pariana belonging to the section with scapose inflorescences, a section found in Amazonia, 
but banished from Eastern Brazil. 

In addition to those above-named, the following is a short list of some remarkable 
Amazonian plants, or plants with Amazonian affinity, met with at Santa Cruz and in 
the neighbourhood : — 

* In fact, no species of tins large tropical genus was seen by me in Matto Grosso. 
t In Martius's ' Flora Brasiliensis,' vol. viii. 
SECOND SERIES. — BOTANY, VOL. IV. 2 Q 



290 ME. SPEXCEE LE M. MOORE — PHANEEOGAMIC BOTANY 



Vismia jajmrensis, Beichardt. 

Psidium insulincola, S. Moore. 

Passiflora coccinea, Aubl. 

Anguria gloriosa, S. Moore. 

Randia Ruiziana, DC.,var. longiflora, K. Schum. 

Faramea bracteata, Benth. 

Salicea novo-granatensis , K. Sclmra. 



Stilpnopappus viridis, Eeiith. 
Tabernamontana hirtula, Mart. 
Maderosperma oblongum, S. Moore. 
Oryctanthus rujicaulis, Eiclil. 
Croton Cajucara, Benth. 
S^eea hermaphrodita, S. Moore. 
Renealmia Holdeni, S. Moore. 



Of the more remarkable Eastern Brazilian plants at Santa Cruz, the subjoined is a 
short list, viz. : — 

Oxalis hirsutissima, Zucc. I Rhodocalyx I'otundifolius, Muell. Arg. 

Hircea sepium, A. Juss. Brosimum Gaudichaudii, Tree. 

Baccharis tridentata, Vahl. i Herreria Salsaparilha, Mart. 

Eut with regard to this list one must remember how little — really almost nothing — is 
known of the South- Amazonian region. This it is which renders doubtful the exclusion 
of any of these plants from Amazonia proper. 

The Primeval Forest Elora. 

It has already been mentioned that the forest extends northward from the immediate 
neighbourhood of Santa Cruz over a gently undulating upland of low elevation as far as 
to the Campos de Tapirapuan. In the gloomy recesses of the forest we should expect 
to meet with a flora different from that of the more open country, so that, instead of 
the latter's xerophilous vegetation, we ought there to be confronted with hygrophilous 
types. Moreover, in passing northward through the forest, we are ever approaching the 
Amazonian plateau. Seeing, therefore, that the North-Brazil-Guiana element is already 
strongly represented in the flora of Santa Cruz, one would a fortiori expect the forest 
flora to show even more affinity with that of the country lying to the nortli. My results 
under this head are somewhat unsatisfactory in consequence of the slender opportunities 
of collecting afforded me during our expedition through the forest, and the consequent 
small number of species secured. It is conceivable, however, that a few specimens, 
plucked hap-hazard may well represent the affinities of a flora ; and in the present case 
the remarkable series of Amazonian species, and species with Amazonian or rather North- 
Brazil-Guiana affinity, can scarcely be a false indication as to the nature of the flora as a 
whole. 

In this flora we hnd a large preponderance of Noi'th-Brazil-Gtuana types, and, as 
compared with the Cuyaba and Santa Cruz floras, marked decrease in the diffused 
Tropical American and East Brazilian factors, as the following ligiir<.>s sliow : — 

The constituents of the forest flora are 

Diifused Tropical American 28 per cent. 

Nortli-Brazil-Giiiana 47 „ ,, 

Common to the two Brazilian provinces 9 „ 

South Brazilian 9 „ ,, 

Mexican 4 „ 



OF TliE MATTO GEOSSO EXPEDITION, 181)1-92. 291 

The reader will also notice here the extremely small continjj^ent of species common to 
the two Brazilian provinces ; so that while at Cuyaha no less than eip^lity-eii^ht per cent., 
and even at Santa Cruz seventy-eight per cent, of the vegetation consists of species one 
might expect to find anywhere in Eastern Brazil, or of plants most nearly related to such 
species, this element of the forest flora amounts to no more than forty-six per cent., the 
proportion of species endemic in regions outside Eastern Brazil being at Cuyaba seventy- 
three, at Santa Cruz eighty-four, and in the forest eighty-eight per cent. 

Here is appended a short list of some forest plants with Amazonian distribution or 
Amazonian affinity ; the distribution or affinity is shown within brackets : — 

Guarea sylvestrls, S. Moore. (Near the N.-]?razil-(iuiana G. paraensis, CDC.) 

Pithecolobium stipulare, Benth, (N.-Brazil-Guiaiia.) 

Bertiera ffuianensis, Anh]. (N.-Brazil-Guiaiia, West Indies.) 

Psychotria triphyUa, Muell. Arg. (N.-Brazil-Guiana.) 

Psychotria homoplastica, S. Moore. (Near P. Ivpulina, Bcutli., from Amazonia.) 

Bletia catenulata, Rniz & Pav. (Guiana, Peru.) 

Rodriguezia secunda, II. B. K. (N. -Brazil-Guiana, Central America, West Indies.) 

Anthurium gracile, Lindl. (N.-Brazil-(juiana, Central America, Trinidad.) 

Anthurium sylvestre, S. Moore. (Near A. Martini, Schott, from (Juiana.) 

Hypolytrum longifolium, Nees. (N. -Brazil-Guiana, Trinidad.) 

My notes contain no record of the finding in the forest of a single plant restricted to 
the South-Brazilian province, and there are only four with exclusive South-Brazilian 
affinity; these are the new species Ilcrpestis parvula, Sohoiimi mittruse, Dalcchampia 
sylvestrls, and Isch}iosipho)i nemorosns. 

The smallness of the collection gives one but slight opportunity to study the ordinal 
peculiarities of this flora. It may, however, be stated that the Bubiacea?, Orchids, and 
Aroids show decided affinity with those of the Xorth-Brazil-Guiana province. Moreover, 
one meets with Zamia Brongniartii, Wedd., on the Tapirapuan campos. The forecjoing 
facts, it is submitted, force upon one the conclusiou that, unless the ei'idence yielded by 
my collection be deceptive, the f ores f-rey ion, although lying icell within the Paraguay 
Valley, must be included within the bounds of the North- Brazil- Guiana, and not, as 
would naturally be sujyposed, within those of the South-Brazilian province. 

Flora of Corumba and Coimbra. 

Corumba is situated close to the ninteenth jiarallel of south latitude, and Coimbra still 
further south. The extreme northern end of the Gran Chaco being not far off, one 
might expect to find, in the flora now to be considered, indications of the proximity of a 
region of comparative drought ; and this, as will soon appear, is to some extent the case. 
The climate of Corumba is much like that of Cuyaba, but I do not know what is 
its annual rainfall. To judge from our own experience, we having spent six weeks 
here in the middle of the wet season, the rainfall is probably less than at Cuyaba. 
Throughout those weeks I do not remember a single day during at least some part of 
which botanizing was impracticable ; and although the district was occasionally visited 
by violent storms, bright sunshine followed by a brilliant sunset was the general rule — 

2q2 



292 ME. SPENCER LE M. MOORE — PHANEEOOAMIC BOTANY 

weather which could scarcely have heen exceptional, as I was informed hy an intelligent 
.old inhabitant, who expressed no surprise at the absence of rain, that Corumba is a " dry 
place." 

The flora of this southern part of Matto Grosso is made up of the following factors : — 

Diffused Tropical American 42 per cent. 

South Brazilian 34 „ „ 

Common to the two Brazilian provinces 15 „ „ 

North-Brazil-Guiana 5 „ „ 

The relatively large number of diffused species will be noticed, as also the great decrease 
in the proportion of types common to the two provinces. The South-Brazilian contingent 
appears in great force, and the jSTorth-Brazil-Guiana is correspondingly diminished. The 
South-Brazilian factor is composed as under : — 

Bras. Or 54 per cent. 

Paraguay (extending into Argentina) 27 „ „ 

Bras. Or., Paraguay, Uruguay 18 „ „ 

And the common element thus : — 

N.-Brazil-Gruiana and Bras. Or 63 per cent. 

N.-Brazil-Guianaj Bras. Or., and Paraguay 26 „ „ 

N.- Brazil-Guiana, Paraguay, and Argentina 10 „ „ 

Amazonian species, or species with Amazonian affinity, met with in the district under 
notice are : — 

Corynostylis pubescens, S. Moore. (This genus is not found in East Brazil, neither did I come across 

it further to the north.) 
Centrosema vexillatum, Benth. (A North-Brazil-Guiana plant.) 
Termnnns voluhilis, Sw. (xVmazonia, Colombia, Central America, West Indies.) 
Tephrosia brevipes, Benth. (Guiana, Colombia, Porto Rico.) 
Pterocarpus Rohi^ii. (N. -Brazil-Guiana.) 

Pacourina edulis, Anbl. (N.-Brazil-Guiana, Colombia, recently found in Paraguay.) 
Macfadyenia laurifolia, Miers. (Venezuela.) 

The Argentine and Paraguayan element is illustrated by such types as : — 



Stiymajihyllon calcaratum, N. E. Br. 
Thinouia sepium, S. Moore. 
Paullinia angusta, N. E. Br. 
Zizyphus oblongifolius , S. Moore. 
Pereskia Bleo, DC. 



Thevetia bicornuta, Muell. Arg. 
Tahebuia Avellanedre, Lorentz. 
Coccoloba sarmentosa, S. Moore. 
Ficus Elliottiana, S. Moore. 
TUlandsia streptocarpa, Baker. 



Adaptations to xerophilous conditions in the form of swollen rootstocks, leathery leaves, 
and stems invested in a thick clothing of cork are frequently met with in this region. 
Several of the genera, such as Cercus^ PeresJda, and Tcdimim, liave fleshy leaves or 
stems. Another interesting xerophilous plant is the noiW Zlzijplins oblong if olius; this 
has the narrow leaves characteristic of species endemic in warm or tem])erate climates, 
and in habit is quite different from tropical species with their reduced spines and large 
leaves ; in fact, the plant is closely allied to one from the dry Argentine country.] 



OF THE MATTO GEOSSO KXPEDITIOX, 1891-92. 293 

ReMAKKS on the PKOOiABLK OllIGIN OF TJIK FOREGOING FlORAS. 

It will scarcely be disputed that tlie whole of Eastern and Central Brazil has, in times 
past, been a great area of evolution and of dispersal. Indeed, tlie large number of 
peculiar genera endemic there — genera, many of them, monotypic or oligotypie — IIk^ wide 
distribution of many of the species, and the high geological antiquity of this part of 
South America are all evidence in favour of the proposition. To tlie student of ])]iyto- 
geography the question of interest here raised is, by wliat means was dis2)ersal elfected ? 
Doubtless the movements of quadrupeds and of birds, wind, dust-storms, &c. have played 
their part in South America as elsewhere, but I cannot help thinking another and even 
more effectual means to have been brought into operation. A glance at the map of 
Brazil will show that the country has two large watcn'-shcds — an eastern, dividing the 
waters of the Sao Erancisco and Parahyba on the east and north-east from those of th<; 
Tocantins on the west ; and a western, running more or less e^st and west, interposed 
between the numerous feeders of tlie Araguaya, Xingu, Tapajos, and Guapore flowing 
northward and westward, and those of the Paraguay with a southward flow. When these 
rivers are in flood many fruits and seeds must certainly l)e carried down to the lower 
country; consequently in the disposition of these water-sheds we can see means for 
ensuring, firstly, Avide disposal of species through Eastern Brazil, and secondly dispersal 
of Central Brazilian types northward to Amazonia, and southward to the country watered 
by the Paraguay and its feeders, such as the Cuyaba, Si])otuba, Caba^al, and Jauru rivers. 
Moreover, from the Silo Erancisco-Tocantins water-shed may Avell have been derived 
species whose distribution is Northern Goyaz and Eastern Brazil; and this may be off'ered 
as a plausible explanation of the considerable resemblance there is between the floras of 
Minas Geraes and Northern Goyaz, although the two districts fall within the bounds of 
diff'erent phytogeographical provinces. Further, the large jiercentage of diffused Tropical 
American types in the floras already dealt with receives a simple elucidation from the 
fluvial geography of Brazil. The density of the vegetation upon the river-banks and its 
comparative sparseness elsewhere have already been mentioned ; this fact must obviously 
enhance the importance of the rivers as distributing agents. One may here refer to the 
occurrence at Corumba, as far south as the nineteenth parallel, of several Amazonian 
plants not found in any immediate station, and, in spite of the strongly South- 
Brazilian character of the Corumba flora, quite unknown in the eastern part of the 
country. What hypothesis sounder than that here advocated can possibly be put 
forward to account for this ? 

But, besides this north and south distribution, there are clear indications as to the range 
of many species of the Matto Grosso flora being dependent upon latitude. This fact will 
be made clear to the reader on cursory examination of the distribution of species 
o-iven in the following pages. He will see that a considerable percentage of East- 
Brazilian types pass through Matto Grosso to Eastern Bolivia, but do not occur to 
the northward. Similarity in the matter of rainfall is probably the reason of this, there 
being a marked difference in that respect between South and North Brazil, and marked 
agreement between the eastern provinces and Matto Grosso. 



294 ME. SPENCEE LE M. MOOEE— PHANEEOGAMIC BOTANY 

I have already shown how greatly the flora of the primeval forest lying hetweeii 
Santa Cruz and Tapirapuan differs from that of the open coimtry extending to the 
eastern bank of the Paraguay river. That such diversity should exist is not difficult to 
understand. The flora of the open country behind the forest (Campos de Tapirapuan), 
so far as my rapid visit enables me to judge, does not show much affinity to that of the 
district round Santa Cruz, although external features are the same in both districts. 
The interposition of a wide tract of forest land is undoubtedly answerable for this diversity. 
In the forest peculiar conditions of light, heat, and moisture prevail — conditions 
unfavourable to most jjlants adapted to the drier, hotter, and sunnier open country. We 
may, therefore, regard the forest as a dense barrier impenetrable from either side. 
Winged, bladdery, and pappose seeds and fruits might occasionally be conveyed by 
winds across this barrier, but this would not be likely to happen often if the strip of 
forest were broad. Berries and drupes, too, would not be distributed by birds, seeing 
how different is the bird-fauna of the forest on the one hand and of open land on the 
other, and for analogous reasons hooked fruits would be little likely to penetrate far. 
On the whole, then, so slight are the means of communication between two open regions 
separated by a broad strip of dense forest, that it would probably not be incorrect to infer 
more floristic dissimilarity under these circumstances than if we had to do with an 
equal breadth of water. Similarly, shade-loving forest plants would only occasionally be 
transported across extensive intervening dry districts freely exposed to the ardours of a 
tropical sun ; and this obvious consideration leads one to believe that almost continuous 
forest must extend from the Upper Amazonian basin to that of the Upper Paraguay, 
most probably via the Guapore river. 

Lastly, a quickly flowing river like the Paraguay at and near Santa Cruz must, in some 
degree, tend to keep neighbouring floras distinct. Plants with edible fruits would indeed 
be disseminated by birds ; but ordinary fruits and seeds, and often also those provided 
with wings or a pappus, &c., would fall into the water to be hurried away down stream. 
At Santa Cruz I was prevented by indisposition from paying much attention to the eastern 
bank of the river, but what little I saw of its flora pointed to decided floristic difference 
between the two river-banks, difference which, at the time, seemed to me accountable 
only in the way just mentioned. Purther observations, however, may prove this imjn-ession 
to be unsupported by facts. 

The collection has been worked up partly at the British Museum, partly at Kew. At 
the former institution the Brazilian flora is illustrated by sets of Gardner, Spruce, Pohl, 
Blanchet, Sello, Poeppig, Von Martins, Bowie and Cunningham, Weir, Claussen, Widgren, 
de Mello, and Messrs. Ridley, Lea, and Bamage. Kew, with most of the above sets, boasts 
the splendid results of the intrepid Burchell's wanderings in the eastern part of the 
country.* A fine set of Biedel's plants, a collection but poorly represented at the 
Museum, is also preserved at Kew ; so, too, are Professor Trail's Amazon specimens, 
and a set of the very large and remarkalde collections made by that en terprisin"* veteran, 

* One cannot but regret that much of the ' Flora Brasiliensis ' should have hcen written without examination of 
British herbaria, especially the two under notice. For this reason, many nondescripts, chiefly of Gardner's and 
Burchell's collecting, exist in this coimtrw 



OF THE MATTO GK0S80 EXPEDITIOX, 1891-92. -295 

M. Glagiou. Moreover, a number of A. St.-llilaire's tjpes have recently been presented 
to the Thames-side institution. I take this opportunity of expressinsj my conviction 
that, when one remembers the comparatively small amount of attention devoted by 
Britishers to the Brazilian flora as contrasted with their eneri^y elsewhere, one has every 
reason to be proud of both our National Herbaria, which, althoujj^h probably not so ricli 
in plants of Brazil as the kindred establishments at Berlin and Paris, do certainly afl'ord 
help quite invaluable to the student. 

It were to shirk a pleasant duty to end tbis introduction without recoi^nition of kind- 
ness received from many quarters in the course of my work. Thanks are especially due 
to Mr. Carruthers, who allowed the plants to be stored at and distributed from the 
British Museum, and who, in every other possible way, has furthered my labours. 
Mr. Carruthers also examined and confirmed my determination in respect of tli(> only 
Cycad in the collection. The other members of the !Museum botaniciil staff have, eaeJi 
in his special department, most obligingly permitted me to take advantage of their 
knowledge. I am indebted to the Officers of the Linnean Society for ki ndly seeing the proofs 
of this memoir through the press in my absence from England. All my friends at Kew 
without exception, from Mr. Baker downwards, were ready with hel]) in cases of difficulty; 
and more than once, when thoroughly beaten owing to the imj)erfect state of a specimen, I 
was put upon the right track by Professor Oliver. Mr. C. B. Clarke very kindly undertook 
the Cyperacea?, an order ui)on which he has bestowed an immense amount of time and 
trouble ; Commelynacese also were determined by the same botanist. Mr. Scott Elliott 
looked at the Eigs, and indicated the novelty and affinity of one of these. Mr. Day don 
Jackson allowed me to consult advance sheets and MSS. relating to his great " Index 
Kewensis," and this saved me much trouble, and prevented some errors. Nor must I omit 
the names of several foreign botanists, who were good enough to examine and compare 
certain specimens and send information on critical points; the list is, I believ(% 
exhausted on mentioning the following : — Professors Engler and B.a(I!kofer, Drs, Schu- 
mann and Taubert, of Berlin, Dr. Warburg, and the late Dr. Morong. To these 
gentlemen niv best thanks are here recorded. 

To Lieut. Olaf Storm and Mr. John Storm I am greatly obliged for doing all in their 
pow er to further my work in every possible way, and also for allowing me to make use 
of the map accompanying this memoir. Dr. Evans, Mr. Holden, and the other 
members of the expedition rendered me much kind assistance, which I here gratefully 
acknowledge. 

I have also to acknowledge indebtedness to the lloyal Meteorological Society, 
through their courteous Assistant Secretary, Mr. William Marriott, for the opportunity of 
access to records relating to the climate of Brazil. Nor should the great pains taken by 
Mr. B. Morgan with the plants figured be left unmentioned. 

Of the plants, the first set, with notes, &c., goes to the British Museum, the second 
■set to Berlin, and the third to Columbia College, l^ew York ; Vienna has the fourth, 
and Kew the fifth. Small sets have also been sent to Paris, Eio de Janeiro, and 
Edinburofh. 



29G ME. SPENCER LE M. MOOEE — PHANEROGAMIC BOTANY 

Systematic Portion. 

DIALYPETAL^. 
THALAMIFLOHM. 

DILLENIACEJi:. 

Davilla lucida, Presl, Reliq. Hoenk. ii. p. 73. 

Sah. Crescit in cacumine montium Serra da Chapada, ubi mens. Aug. floret. (N. 109.) 
[Anier. Trop., Ind. Occ] 

Davilla lacu^^osa, Mart, in Plora, xxi. (I808), II. Beibl. p. 49. 

Kah. Matte Grosso (Leeson (n. 7) in Herb. Brit. Mus.). [Bras. Or.] 

CuRATELLA AMERICANA, L. Syst. ed. X. 1079 ; Sp. PI. ed. II. 718. 

Hah. Reperi inter Cuyaba et Serra da Cbapada mens. Aug. florentem, itaque ad 
Santa Cruz, ubi Llcha nuncupatur. (Nn. 54, 423 a.) [Amer. Trop., Ind. Occ] 

DoLiocARPiJS DENTOsus, Mart, in Plora, xxiv. (1811), II. Beibl. p. 65. 
Hah. Ad Santa Cruz flores prsebet mens. No\r. (N. 535.) [Bras. Or.] 

ANONACE^. 

Ephedeanthus, e tribu Uvariearum gen. nov. 
(Plate XXI.) 
Elores actinomorphi. Sepala 3, aestivatione valvata. Petala 6, libera, biseriata, omnia 
superne imbricata, inferne aperta, subsequalia, erecto-patentia. Stamina indetiuita, 
dense pluriseriatim imbricata, cuneata, a dorso compressa; connectivus sursum 
dilatatus, incrassatus, truncatus ; antherse extrorsae. Torus hemisphtericus. Frutex 
diffusus, ramosus. Eolia disticba, penninervia. Elores pusilli, unisexuales (an dioici?), 
feminei ignoti, masculi extus sericeo-tomentosi, in axillis foliorum solitarii, brevissirae 
pedunculati ; pedunculi bracteis latis 3-seriatis distichis, arete imbricatis, alabastra 
obtegentibus onusti. Carpella et baccse ignotse. 

Ephedranthus parvielorijs, S. Moore ; ramulis teretibus pubescentibus, demuni 
subobsolete puberulis ; foliis breviterpetiolatis, ovato-lanceolatis vel obovato-oblongis, 
obtusis acutisve, interdum breviter cuspidatis, basi gradatim angustatis ibique sub- 
rotundatis, rarius levissime cordatis, tenuiter coriaceis., supra glabris, subtus, prtescr- 
tim in nervis, appresse pubescentibus, dein glabris ; costis secundariis utroquc latere 
6-9, adjectis pluribus tertii ordinis, angulis variis, plerumque vero latis, insertis, primo 
nunc leviter, nunc levissime arcuatis, deinde undulatim-fornicatira-conjunctis, supra 
impressis, subtus eminentibus ; pedicellis ferrugineo-tomentosis, bracteis absconditis ; 
bracteis rotundatis vel rotundato-ovatis, obtusis obtusissimisve, extus ferrugineo- 
sericeis, intus levibus, puberulis ; sepaiis rotundato-ovatis, obtusis ; petalis sepala 
paullo excedentibus, sericeo-tomentosis, exterioribus ovato-oblongis, quam interiora 



OF THE MATTO GEOySO EXPEDITION, 1891-92. 297 

oblonga paullulum brevioribus, omnibus obtusis; stamiiiibus circa S-seriatis })arvis, 
subsessilibus, antherarum lociilis basi obliquis. 
Mab. Crescit in sylvulis juxta Santa Cruz, ubi flores pra'bet mens. Sept. (N. 310.) 
Eamuli usque O'l cm. diam., ultimi vero vix 0*1 cm., fusco-grisei, loniijitrorsum rimosi, 
minute lenticelliferi. Polia 55 110 cm. long., 2-0- 1-5 cm. lat., supm vix nitidula, 
subtiis paullo pallidiora. Pedunculi 0*2 cm. long. Bractea? basi late inserts, infimse 
exiguse, superiores majores, ilL-E equidem 0-08 cm. long., intermedige infimas plus 
quam duplo oxcedentcs, summEe 0-35 cm. long., omnes coi'iacost\ Floros circa 
08 cm. diam., dilute viridcs. Sepala vix O'l cm. long., 0-37 cm. lat., coriacca, extus 
ferrugineo-sericea, iutus appresse puborula et saltern in sicco cnstanea. Petala 
submembranacea, exteriora 0*5 cm. longa, omnia intus plus minus pulxMiiLi. Torus 
O'l cm. alt. Stamina modo 0*07 cm. long. ; antberarum loculi O'Ot cm. long. 
Genus verisimiliter juxta Gnatfcriam intcrponendum, abs qua optime abborret ilorilvus 
unisexualibus jmrvulis subsessilibus bracteis obtectis, pctalis omnibus manii'estc imbri- 
catis nee unquam subvalvatis, et toro bemispbicrico. 

This plant, on account of its aistivation and the structure of its androDcium, is undoubt- 
edly to be referred to the tribe Uvaria\ Probably it will come nearest to Guatteria, but 
this will depend on the nature of the placentation. Should the carpels prove to be 
many-ovulate, Ephedranflms will have to be placed in the neighbour! lood of the East 
Indian Sagercea and StelechocarpuSt the former of which may have unisexual flowers, 
while the latter is dioecious. 

I have examined all the flowers it was possible to sacrifice, in order to ascertain 
whether carpels are present, but in no case could a trace of one be found. The floAvers, 
therefore, if they be not dioecious, are at least unisexual ; for myseK I strongly incline 
to the former supposition. Upon either view, the genus would be markedly different, 
exclusive of other characters, from Guatteria^ the flowers of which are always 
hermaphrodite. 

Here follow a few notes upon various points of interest possessed by the plant : — 
The stem * (Plate XXXVIII. fig. 1) is in no way abnormal as respects its structure. 
The protoxylem is situated at the extremity of each mass of xylem, which projects some 
distance into the pith ; in its immediate neighbourhood are a number of small fibres and 
tracheides, and these latter are succeeded further outwards by pitted tracheides and fibres 
of ordinary appearance ; the xylem parencbyme, which is comparatively rare, shows a 
tendency to arrange itself in tangential lines. The soft bast is normal, while the hard bast 
in each phloem region is composed of two or three masses of fibres, of which the outer- 
most has the greatest thickness. These masses frequently run right across and so stop up 
the medullary rays. There is a narrow pericycle ; it is bounded externally by a layer, in 
some places double, of thick- walled cells forming a well-marked endoderm. The cortex, 
especially its outer portion, is strengthened by means of frequent thick-walled sclerotic 
parenchyme cells, a little way outside which, and close to the epiderm, is the phellogen 

* Here and elsewhere in this part of my report on the botany of our expedition only a few of the more obvious 
facts of structure have been recorded. 

SECOND SEEIES. — BOTANY, VOL. IV. 2 K * 



298 ME. SPEXCEE LE M. MOOEE— PHANEEOGAMIC BOTANY 

layer. The medullary rays vary considerably in width; like the cells of the com- 
paratively small pith-region, they are filled with spherical or elliptical simple or more 
usually compound starch-grains, varying greatly in size, the largest being usually found 
in the pith. 

The leaves are inserted on a strictly distichous plan, but material for the study of the 
arrangements of the leaf-traces is not available. The one point worthy of mention a 
propos of the leaf- structure is that the stereome protecting the vascular bundles runs 
right up to the epiderm, which latter tissue, where it overlies the bundles, is composed 
of smaller cells than usual, and of cells often with somewhat thicker walls than else- 
where ; in other words, the epiderm itself can to a certain extent assume the appearance 
of stereome, and this arrangement must obviously give great strength to the leaf. 
Beneath the midrib, however (that is, on the lower side of the leaf), parenchyme inter- 
venes between the stereome and the epiderm (fig. 3). Pig. 2 shows the upper surface of 
the leaf seen from above ; over the palisade tissue the epiderm cells are larger and have 
the wavy borders so frequently possessed by epiderm; but the cells lying over the vascular 
bundles are shorter and more angular, and usually, though not quite always, each of 
them contains a single rhomb oidal crystal. The leaves are bifacial in the most marked 
degree, there being no stomates on the upper surface ; each stomate is accompanied by 
two subsidiary cells, often very unequal in size (fig. 4). Near the vascular bundles the 
mesophyll seems to be continuous, without any intercellular spaces ; further away, 
however, it becomes " spongy." Many Anonaceaj have glandular leaves, and 
Uphedranthus is no exception to this, as sections show several glands immersed in the 
lower part of the mesophyll. These glands are similar in appearance to those of 
Stormia, shown in fig. 7 of Plate XXXVIII. 

Like the leaves, the three series of bracts are strictly distichous. In the case of one 
fliower, however, which I examined, the sixth bract did not stand exactly over the fourth, 
but had been deflected to a trifling extent towards one side, probably as a result of 
abnormal pressure in the bud. In the diagram (Plate XXXVIII. fig. 5) it will be seen 
that the position of the axis is not indicated ; this I have not been able to ascertain, 
owing to want of material. 

The small pollen-grains are ellipsoidal in shape, and, except for a few fine striations, 
are quite smooth. I could make out two pores, but never more. 

GuATTERiA SYLVICOLA (sp. nov.) ; caule subtereti, leviter undulato, nigrescenti-purpur- 
ascente rufo-tomentoso, dein pubescente, demum glabro ; foliis maxime in^qualibus, 
subsessilibus, elongatis, anguste oblongo-obovatis, cuspidatis, obtusis, basi cuneatim 
angustatis, coriaceis, supra fere glabris, subtus, prsesertim secus nervos, rufo-tomentellis ; 
peduncuhs solitariis, foliis longioribus, prope basin articulatis, indumento rufo-villoso- 
tomentoso instructis ; bracteis rainutis ; sepalis late ovatis, basi pauUo cordatis, 
obtusiusculis ; petalis obovato-oblongis, obtusis vel retusis vel brevissime bilobis, 
sepala duplo excedentibus ; staminibus late linearibus obtusissimis ; baccis ignotis. 
Sab. In sylvis ad Serra da Chapada, alt. circiter 2000 ped. supra mare PI Auo". 

(N. 142.) 



or THE MATTO GKOSSO EXPEDITION, 1891-92. 2m 

Polia IO'O-27'O cm. long., medio 3-5-8-0 cm. lat. Arbuscula liahitu fruticis elati. 
Caulis undulatus, domum reticiilato-insculptus, 0-4 cm. diam., margiiie imdulato ; 
costxe secundariie utrovis latere 15-20, adjectis paucis tertii ordiiiis, angulis latis 
alternatim insertis, raro suboppositae, rectic, infra marginem subito forniealim con- 
junctse, costis pag. sup. eminentibus, nervis laxe recticulatis, paruni i)roniinulis ; 
petiolus 0-5 cm. long,, incrassatus, late canaliculatus, rufo-tomentosus. Bractese 
ovata?, acutsB, tomentosee, 0*2 cm. long. Pedunculi 1-0-1-5 cm. long., erecti. Sepala et 
petala crassa, ilia extus rufo-tomentosa, 0-8 cm. long, et lat., Iiaec I'O cm. long., apice 
vix I'O cm. lat., panllo supra basin ad 0-6 cm. angustat<i, juxta basin rufo-, superne 
cinereo-tomentosa. Torus stamineus O'll cm. alt. Stamina C-seriata, circiter 
0*12 cm. long. Gynaecium 0-3 cm. diam., 0*22 cm. alt., medio levissime elevatum ; 
carpella tetragona, rufo-tomentosa, cum stigmate prominulo 0'2 cm. longo. 
Videtur prope O. Ouregou^ Dun., pononda, sed diversa, pra'ter alia, ob folia majora 

et comparate angustiora, basi cuneatim angustata, pedunculos solitjirios loliis 

breviores, &c. [Amazonia^ 

DrGUETiA FURFUiiACEA, Bcuth. & Hook. f. Gen. PI. i. p. 24. 

Hah. Crescit in cacumine montium Serra da Cliapada, ubi fructificantem habui mens. 
Aug. (N. 163.) [Bras. Or.] 

DuGUETiA Sanct^-crucis (sp. uov.) ; caule tereti, in sicco rimoso, Icvitcr fiirfuracco, 
deinde fere glabro et suffusccnte ; foliis coriaceis, breviter petiolatis, elongatis' 
lanceolatis, obtusis, acutis, vel cuspidato-acuminatis, supra nitentibus, subtus 
arctissime furfuraceo-tomentosis, neryis subtus eminentibus ; pedunculis solitariis, 
erectis, petiolos longe excedentibus, juxta medium articulatis; bracteis parvis, 
rotundatis, obtusis ; ovariis linearibus, apice in stylum gracilem rectum vel incurvum 
et ovario duplo breviorem desinentibus ; carpidiis (vix maturis) obovoideo-oblongis, 
angularibus, sericeo-fusco-tomeutosis, basi aspermis, rostro recto vel plerumque 
incurvo coronatis. 
Hah. Crescit in sylvaticis ad Santa Cruz, mens. Nov. florens. (N. 576.) 

Arbor parvus, circa 6 met. alt., habitu fruticoso. Caulis ascendens, sat crebro ramosus. 
Petiolus 0"4-0-5 cm. long., sulcatus, incrassatus, arete annulatim striatus, furfuraceo- 
tomentellus ; lamina 9-0-21'0 cm. long., 2-5-5-2 cm. lat., basi angustata vel levissime 
rotundata, utrinque reticulato-nervosa ; costse secundariae utrovis latere 12-16, 
adjectis paucis tertii ordinis, angulis variis latis insertse, primo rectae vel undulatae, 
dein dichotome f ractae et antrorsum et retrorsum arcuatim porrectae, nunc simpliciter 
ac magnopere arcuatae. Bractea? cirai 0-5 cm. long., ami^lexicaules. Pedunculi 
2-0 cm. (sub fructu attamen 35 cm.) longi, superne leviter incrassati, furfuraceo- 
tomentosi. Sepala, petala, et stamina desunt. Torus stamineus 0-15 cm. alt. 
Gyneecium 0-45 cm. diam. Ovaria 0-22 cm. long., dorso compressa. Eructus 
immaturus fere 2-5 cm. diam. Carpidium basi carinatum 0-6-1-0 cm. long., 
basis asperma compressa, 0-2-0-6 cm. long. ; rostrum 0-1-0-2 cm. long., subulatum, 

dorso compressum. 

2r2 



300 ME. SPENCEll LE M. MOORE — PHANEROGAMIC BOTANY 

Species propria ob pedunculos bracteatos forsan ad D. hracteosam, Mart., accedens, ab 
ea vero aifatim distat. [Bras. Or.] 

Stohmia, e tribu Unonearum gen. nov. 
(Plate XXII.) 
Plores actinomorphi, bermaphroditi. Sepala 3, reduplicato-valvata, crasse coriacea. 
Petala 6, biseriata, testivatione valvata, in corollam gamopetalam interne connata, 
interiora quam exteriora paullo minora, exteriora a calyce paullulum superata. 
Stamina ao , pluiseriata, arete imbricata, anguste cuneata, connectivo apice truncato- 
dilatato, loculis transverse rugatis, uniseriatim plurilocularibus, extrorsis. Torns 
parum elevatus, apice late truncatus. Carpella 14*-21, sessilia, circa 7-ovulata, stylo 
infundibular! majusculo incrassato elongato integro coronata ; ovula sub-1-seriata, 
anatro2)a. 33acc8e (ex St.-Hil. & Tul.) " septis spnriis transversim multiloculares." 
Prutex altns, vel potius arbuscula. Eolia breviter petiolata, costis 'secundariis 
plurimis percursa. FJores mediocres, solitarii, pedunculis exaxillaribus insidentes. 
Receptaculum dilatatum, laminatum, paullo ultra sepalorum insertionem annulatim 
productum. 

Stoemia BRASiLiENSis, S. Moore. — Hexalohus braslllensls, A. St.-Hil. & Tul. in Ann. Sc. 

Nat. ser. II. xvii. p. 133. — Trigyneia brasiliensis, Benth. & Hook. f. Gen. PL i. p. 24i. — 

TJnona hrasiliensis, Baill. Adansonia, viii. p. 305. Slirps fere 4-metralis, ramosa. 

Bami abundanter foliati, subteretes, arete tomentelli, nunc pubescentes, deinde fere 

glabri et rimis anguste ellipticis notati, fusci. Polia oblongo-lanceolata vel oblocu- 

lato-oblonga, breviter cuspidata, obtusa, rare obtusissima vel emarginata vel retusa, 

basi ssepissime paullo obliqua angustata vel subrotundata, cito costa media appresse 

tomentella exempta glabra, nitida ; costse secundariae utrinque 15-20, angulis fere 

rectis paullulum decurrentim insertac, juxta marginem subito arcuatim-undulatim 

conjunctse, utrinque prominulse ; venulai laxe reticulata^, utrinque eminentes. 

Petioli circa I'O cm. long., aliquatenus complanati, tomentelli. Pedunculi circa 

I'O cm. long., plerumque plus minus decurvi, arete albido-sericei. Plores 

pallide virides. Alabastra trigona, paullulum ante floritionem circa 10 cm. 

diam. Sepala ovata, obtusa, extus albido-sericeo intus tomento dilute ocbraceo 

munita, 12 cm. long., 0'8 cm. lat. Petala ovata, obtusa, interdum iugequalia; 

lamina libera, basi leviter cordata, albido-sericea, intus in sicco fusca, Beceptaculum 

0'4 cm. diam.,aliquanto triangulare. Torus circa 0*15 cm. alt., apice 0-25 cm. diam. 

Stamina 0'3 cm. long. ; filamenta brevissima ; connectivus sursum gradatim 

amplifiaitus, apice parum incrassatus. Carpella lincaria, albo-sericea, arete conferta, 

0"2 cm. long. ; styli ovarium semieequantes, apice pilis stigmatosis instructi. Baccse 

non suppetebant. 

Ilab. Ad Santa Cruz, floret mens. Nov. (N. 523.) 

Genus ad Vnonam, necnon ad Aswiinam, spectans. Ab illii attamen divcrgit imprimis 
receptaculo dilatato, petalis parvis connatis, et staminibus rugatis ; ha?c, contra, reeep- 
taculo normali, petalis liberis, toro subgloboso, staminibus et stigmatibus disparibus, 
ovulis biseriatis gaudet. Cum Trigyneia, cujus flores pusilli, receptaculum baud 



OF THE MATTO GHOSSO EXPEDITIOX, 1801-92. 301 

dilatatum, et petala libera, et carjiella indofinita, certe non congcnericji. He.vaJobiis 
tametsi petala Lasi connata, itaque flores ])arvos, roce])taculum normalo, stigmata 
bipartita, carpella pluriovulata plerumquc biseriata i)r8ebet. 

Undoubtedly tbe most noteworthy feature about Sformia is the curious dilated 
receptacle. On turning the flower upside down this ajipears as a membranous obscurely 
triangular expansioi. of tbe top of tbe peduncle, the sepals being inserted well within its 
free rim. In this respect Sformia holds to other Anonacea? the position o^ EachschoUzm 
to normal Papaveraoeae, but the comparison nuist not be carried too far, as there is not 
the least sign, in the Anonacea^, of that perigyny Avhich is so remarkable a feature in 
Eschscholtzia. Curiously enough, authors have comi)letely overlooked this dilat«ation of 
the receptacle, although it is somewhat obscurely figured by St.-Hilaire and Tulasne. 
The Australian Euimmatia, it may be observed, also has a dilated receptacle ; but, inas- 
much as the carpels are immersed in it, no valid comparison of that genus with Slormia 
is possible. I cannot endorse Prof. Baillon's remarks (' Adansonia ' viii. ]). 300) : — 
"MM. Bentham et Hooker ont tres-sagement rapj^orte au genre Trlgyneia \UexalohuH 
hrasiUensis, A. St.-Hil. et Tul." Baillon's views are very original, and if they were accepted 
the genera of AnonaccLC would be greatly reduced in number. Thus, he sinks not only 
Trigynekif but Melodorum and Cananga also in TJnoHa, a genus which he regards as 
embracing no less than fifteen sections, all founded on geiu^ra of former authors. 
Moreover, Baillon, speaking of our plant, says (' llistoiredes Plantes,' i. p. 212) : — " Sauf 
I'union de ses petales, cette espece est tres-voisine de notre Unona (Hlrcriana-,^^ and this 
leads one to doubt whether the Paris Professor could have examined the Brazilian type. 

The structure of the curious corrugated anthers of Slormia is alludc^d to further on. 
It must suffice here to point out their resemblance t<j those of that singular genus 
Hornschnchia of Nees von Eseubeck. This genus Xees at first referred to Sapotaceae, 
soon removing it to Olacinese, and afterwards to Sapindacete. Endlicher placed it 
doubtfully at the end of Ebenaceae ; Lindley, also Avith doubt, among the Sapindaceae ; 
Meissner considered it to be a Sapotacea, while ^liquel, after thorough examination, 
suggested an entirely new afiinity, viz., with Lardizabalese. In the ' Genera Plantiirum,' 
on the other hand, Bentham and Hooker exi)rcss their belief in its being Anonaceous. 
Here is disagreement among the doctors indeed ! The description of the flower is briefly 
as follows : — There is a deep cup-shaped outer organ, presumed to be a calyx ; the petals 
are 6, biseriate and valvate in aestivation ; stixmens are inserted on a slightly' raised 
torus, and the anther-cells are transversely corrugated, and divided each into a longitudinal 
series of chambers ; the pollen is compressed. There are three free carpels, each with 4 
or 5 uniseriate anatropous ovules, and each crowned with a small stigma. The fruit is 
A berry, to some extent spuriously septated ; the ripe seed, unfortunately, is not known. 
"Were the receptacle of Stormia to become deepened, we should ^et something like the 
calyx of Hornschnchia. But I suppose the latter organ must be a true ciilyx; otherwise, 
on the hypothesis of afiinity between Sornschiwhia and Stormia , what has become of one 
of the three outer whorls of its floral leaves ? The paucity in stamens, when contrasted with 
Stormia' s multiple andrcEcium, is a striking point of difference between the two genera ; 
this, however, seems over-ridden by the remarkable septation of the anther-cells in both. 



302 ME. SPENCER LE M. MOORE — PHANEROaAMIC BOTANY 

The pollen of Hornschuchia takes the form of large multicellular m.asses ; that of 
Stormia, though often consisting of single grains, is frequently found, even when 
obtained from anthers apparently adult, to be more or less aggregated owing to per- 
sistence of the special anther-cell's walls (Plate XXXVIII. fig. 12). Except for the 
numlier of its carpels, the gynsecium of Hornschuchia well stands comparison with that of 
Stormia, and the septation of the berry is another point of agreement. Bearing all these 
facts in mind, I certainly think the above-stated view of Bentham and Hooker to be that 
one for which there seems to be the most warrant. 

In the structure of its stem (Plate XXXVIII. fig. 5) Stormia presents no point of 
special interest. There is a relatively reduced pith, of which the cells are filled with 
small spheroidal or ellipsoidal starch-grains, sometimes simple, sometimes compound. In 
the xylem there are numerous (about twenty) masses of protoxylem, and tracheides are 
fairly abundant. Pibres of normal appearance make up most of the remaining part of 
the xylem, paren chyme-cells being rare. In the phloem the fibres, parenchyme, and small 
sieve-tubes with companion-cells are in every respect normal. The sections made by me 
showed usually four groups of bast-fibres ; of these the three inner have their long axes 
tangential, the outer group being elongated in a radial plane. There is a well-marked 
endoderm. The medullary rays, one, two, or three cells thick in the xylem, become 
widened on passing into the phloem, and are here sometimes strengthened by single or 
aggregated sclerotic parenchyme elements ; similar elements are also to be seen in both 
cortex and pith. The phellogen layer lies close to the epidei'm. 

The venation of the leaf is peculiar and reminds one more of a Ckisia than of an 
Anonacea. The peculiarity is mainly owing to the presence of three nerves of tertiary 
order in the interspace between a secondary nerve and its successor. These nerves run 
more or less parallel with the secondary nerves, and they may anastomose with these or 
with each other. There are no stomata on the upper surface of the leaf, and the 
epidermal cells here contain each a single spheraphidal mass. The palisade parenchyme 
is one cell deep, the layer immediately below it having it cellss somewhat elongated 
perpendicularly. But few air-spaces occur in the spongy mesophyll, and even these are 
small. Immersed in the lower part of this tissue are numerous glands of normal 
appearance. The cells of the lower epiderm usually contain one small mass of 
spheraphides ; rarely two such masses are seen, or a prismatic crystal takes the place 
of the spheraphides. Stomates are numerous in this layer, each stomate having a couple 
of subsidiary cells (fig. 6) . 

The arrangement of the leaf-traces is simple. The three bundles of the petiole unite at 
the axil into a single strong common bundle which courses down through two internodes, 
and forks at the second node below that at which it entered the stem. Each of these 
forks is markedly thinner than the parent bundle, and the shorter of them inserts 
itself on the neighbouring trace of its own side close to the node, while the longer runs 
some distance down the third internode to finally unite with the same trace. 

By referring to figs. 9-12 of Plate XXXVIII., the reader will at once understand the 
structure of the anther. As seen in transverse section, the greater part of this organ consists 
of the relatively large connective, composed mainly of a plexus of delicate cells with brown 



OF THE MATTO GEOSSO EXPEDITIOX, 1891-92. 303 

walls, in wliicli are immersed numerous sclerencliyme-eells of a yellowish colour ; this 
tissue is traversed by the exceedingly slender vascular bundle. Outside tliis come two 
layers of larger cells A\'ith light brown walls, and of these the outer abuts on the delicate 
epiderm. At each end of these two masses of light brown colls is a group of colourless 
cells with somewhat thickened walls. The exotheeium and endotlieciuni are l)oth norma], 
and, even in the case of anthers from dried flowers, which are all that I have been able to 
examine, remains of the more or less disorganized tapetum can be made out. 

As has already been explained, the anther is multitubular, with loculi arranged one 
beneath the other in Hnear series to the number of about fifteen (figs. S & 9). Between 
each pair of loculi is interposed a delicate septum, of which the structure is difUcult to 
make out from dried anthers; each in surface view is seen to be composed of exceedingly 
thin-walled cells, decreasing in size from periphery to centre. When the development of 
these anthers comes to be worked out, I suppose it will be found that the archesporial 
cells, instead of forming a continuous series, are interrupted by thin strips of indifferent 
tissue. The anther-cells open in the usual way, beginning from above downward, and it 
would appear that the septa are ruptured during the i)rocess ; at any rate, in anthers 
from which the pollen has fallen I could see no trace of septa. A\'itli the exception of 
Hornschuchia^ I know of no anthers with these serial loculi ; they recall to a very slight 
extent the anthers of the Mimoseae. 

HoLLiNiA INCURVA (sp. noY.); foliis brevipetiolatis, lanceolatis vel lanccolato-oblongis, 
obtusis, basirotundatis, coriaceis, supra nitidis velnitidulis, glabris, subtuspallidioribus, 
minute puberulis ; costis secundariis utrinque circa 12, angulis latis insertis, sub 
margine undulatim et fornicatim conjunctis; pedunculis solitariis bigeminisve, 
petioles excedentibus, basi bracteatis, et juxta vel supra, nonnunquam vcro infra 
medium bracteoM parva onustis, puberulis; floribus modicis, calyce forrugineo- 
pubescente, coroUse arete ferrugineo-tomentosse alls spathulato-oblongis ascendenti- 
incurvis ; fructu ignoto. 
Hab. Crescit in sylvestribus ad Santa Cruz, ubi mens. Nov. floret. (N. 522.) 

Trutex diffusus longiramosus, vix usque trimetralis. Ramuli teretes, purpurascentes 
vel purpureo-grisei, leviter undulati, creberrime longitrorsum rimati, juveniles 
puberuli. Poliorura lamina plerumque 100-14-0 cm. long., et 3-5-1-5 cm. lat., 
deorsum paruni obliqua ; costa media supra impressa, subtus admodum eminens ; 
costa? secundarite utrinque prominulai ; venulse arete reticulata^, supra maxime 
aspectabiles ; petioli 0-8-1-0 cm. long., supra anguste canaliculati, puberuli. 
Pedunculi usque 30 cm. long., superne amjiliati. Bracteac bractcolaeque ovatae, 
acut^, extus ferrugineo-tomentellse, circa 0-15-0-2 cm. long. Elores lutescenti- 
virides. Calycis lobi abbreviati, rotundati, carnosuli, concavi, dorso carinati, 
0-25 cm. long. Corollas basi 07 cm. diam., alae usque 1*5 cm. long., superne 
0-65 cm. lat. Antherae O'l cm. long. Torus stamineus 0-2 cm. alt., 0'25 cm. 
diam., pubescens. Carpella albo-sericea, 0-15 cm. long. 
Ex affinitate B. orthopetahe, A. DC, arbor cujus folia diversiformia, et petala majora 

una cum calyce canescenti-tomentosa. [Amazonia, Guiana.] 



304 ME. SPEK^CEE LE M. MOOEE— PHAISEEOGAMIC BOTAIVY 

Akona coriacea. Mart. Fl. Bras. xiii. I. p. 6, ex icon, et descript. 

, var. AMPLEXiCAULis (S. Moore). A typo clistat imprimis foliis minoribus insigniter 

amplexicaulibus. Forsan pro specie propria existimanda. 

Sah. Floret ad Santa Cruz mens. Nov. ArrlcMcum incolariim. (N. 517.) 

Petala admodum incrassata, intus liiteola. 

E-hizomatis fimbriae ad vnlnera a morsnbus serpentum facta applicantur. 

In habit our plant, which is frequent at Santa Cruz, is apparently different froui 
typical A. coriacea, which Martins describes as having tortuous branches, inasmuch as it 
is a lanky erect shrub, 4-6 feet high, and very sparingly branched, if at all. 

The popular name " Arrichicum " is obviously merely a local variation of the Araticu 
of East Brazil, a term applied to several species of Anona there native. [Bras. Or.] 

Anoxa dioica, a. St.-Hil. Fl. Bras. Merid. i. p. 34 (ex descript.). 

Hah. Yiget in locis apertis arenosis ad Santa Cruz, mens. Nov. florens. (N. 693.) 
Eandem plantam legerunt beat. Gardner ad Santa Rosa in provincia Pernarabuco 
(n. 2774 in hbb. Brit. Mus. et Kew) et beat. Burchell (n. G390 in Hb. Kew). Crescit 
itaque in Paraguaria, unde misit cl. Balansa (n. 2298). [Bras. Or., Paraguaria.] 

Akona (§ GuANABANi) Walkeri (sp. nov.) ; caule abunde ramoso, ramulis subteretibus, 
juvenilibus rufulo-tomentosis, mox glabris, sparsim lenticelliferis ; foliis brevipetiolatis, 
parvis, maxime insequalibus, ovatis, obtusissimis vel breviter apiculatis, basi pauUo 
rotundatis, interdum leviter coarctatis, supra puberulis^ subtus piloso-pubescentibus, 
membranaceo-coriaceis, obscure pellucide punctatis ; pedunculis folia suba^quantibus 
vel quam ea brcAdoribus, arete rufo-tomentosis ; basin versus minute 1-bracteatis ; 
bracteis subulato-linearibus, rufo-tomentosis ; sepalis parvis late triangularibus, longe 
acuminatis, rufo-tomentosis; petalis exterioribus late ovatis, pet. interioribus 
quam exteriores paullo brevioribus lineari-oblongis, omnibus acutiusculis, arete 
velutinis ; gynsecio depresse subgloboso rufo-velutino ; fructu ignoto. 
Hah. Crescit juxta urb. Cuyaba, ubi Acres mens. Aug. prsebet. (N. 38.) 

Frutex erectus, circa 3 met. alt. Bamuli fusco-cinerei, saltern, in sicco crebro et alte 
rimati. Folia majora cum minoribus mixta, 2*4-5*8 cm. long., medio 2'3-3'6 cm. 
lat., subtus pallidiora ibidemque eximie nervosa ; costge secundarise utrinque 6-8, 
levissime curvatai, marginem versus dichotomy; indeque arcuatim conjunctse; 
venuiae laxe reticulata?, parum prominulse; petioli non ultra 0*3 cm. long., 
canaliculati, basi ampliati, rufo-tomentosi. Pedunculi ad 3*2 cm. long., medio 
0*13 cm. crass., erecti, sub flore sensim incrassati, usque ad 08 cm. a basi bracteati. 
Bractea; circa 0'25 cm. long., pedunculo appressse. Flores vix 3 cm. diam. 
Sepala viridia, vix 1-0 cm. long., basi O'o cm. lat., eorum acumen 0*7 cm. long. 
Petala exteriora 1*4 cm. long., medio 1*2 cm. lat., pet. interiora 1-2 cm. long., nee 
ultra 0'7 cm. lat. ; pet. omnia carnosa, extus lutea, intus lactea, et secus medium 
purpurea. Stamina 0*2 cm. long. Gynsecium 0-2 cm. alt., 03 cm. diam. 
Videtur cum A. cornifoJia, A. St.-Hil., et A. spincsceuti, Mart., comparanda, ab ilia 

abhorrens statiira altiore, foliis minoribus, pedicellis nee supra medium bracteatis, sepalis 



OF THE MATTO GKOSSO EXPEDITION, 1891-92. 305- 

longe acuminatis, ab liac vero ob spinarum abscntiam, folioniin indumentum, pedicellos 
1-bracteatos dispar, et, alias inter notas, sepalis omnino diversis facile dii;-iio.seenda. 

The species has been named after Mr. J. 0. Walker, a gentleman interested in botany, 
whose acquaintance I made at Cuyaba, and who very kindly accompanied and helped me 
while collecting near the city. [Bras. Or.] 

Anona (§ Att^ ; Pil^flor.e) Sanct.e-crucis (sp. nov.) ; caule erecto, ramoso, ramulis 
teretibus, rufo-pubescentibus, dein fere glabris,paUide bruncis,abimde necnon grosse 
lenticelliferis, in longitudinem crebre rimosis; foliis brevipetiolatis, oblongis vel 
obovato-oblongis, obtusissimis, basi coarctatis, supra fere glabris, subtus, praisertim 
in nervo medio, minute rufo-pubescentibus, coriaceo-mcmbranaccis, crebre ac 
minute punctatis ; pedunculis quam petioli longioribus, crassis, juxta basin 
1-bracteatis, sericeo-velutinis; bracteis late ovatis, crini brevissiino coronatis; s{^i)alis 
deltoideo-ovatis, obtusiusculis, arete velutinis ; petalis exterioribus ovato-rotundatis, 
obtusis, arete rufulo-velutinis ; gynsecio subgloboso, tomentoso ; fructu sii])globoso, 
superne paullo coarctato, sparsini ac brevissime tuberculato. 
Hah. In nemorosis ad Santa Cruz (Barra dos Bugres), ubi tioret mens. Nov. et mens, 
insequente fructum praebet. Nom. vernac. Ata. (N. 585.) 

Arbor parva, circa 6 met. alt., erecta, ramos crebros patulos elongatos gerens. Folia ad 
15*5 cm. long, (pleraque vero circa 12*0 cm.}, medio i-7-C-5 cm. lat., marginc 
levissime undulata, utrinque pulchre reticulato-nervosa ; nervus medius supra 
impressus, subtus valde eminens ; petiolus crassus, puberulus, saltern in sicco arete 
transversim rimatus, 0*7 cm. long., 0*2 cm. crass., late canaliculatus. Pedunculus 
vix 1'6 cm. long, (sub fructu vero 3'5 cm.), medio 0'25 cm. crass, (sub ilore ipso vix 
0'4 cm., et sub fructu prorsus 0*4 cm.). Bractece 0*4^ cm. long., 0*3 cm. lat., firmse, 
velutinge, margine ciliolatse, crebre punctatae. Flores virides, forsan circa 2"5 cm. 
diam. Sepala 0"5 cm. long., 0*6 cm. lat., firma, arete albo-ciliata, punctata. Petala 
(vix matura) circa 1"5 cm. long., et medio 1-6 cm. lat., carnosa, utrinque velutina ; 
petala interiora nobis non obvia. Gynaecium floris vix maturi circa 15 cm. diam. 
Fructus (eduhs et ab incolis magni laudatus) fere maturus magnitudine eam 
Citri aurantii aequans, 6'0 cm. long., medio 0*0 cm. lat. (sub apice modo 3-0 cm.), 
viridis, tuberculis nee ultra O'l cm. e superficie minute scrobiculata eminentibus 
(nonnunquam vero evanidis) sparsim obtectus. Semina 1*3- 1"4 cm. long., medio 
0*9 cm. lat., ad hilum Oo cm. lat. ; testa Crustacea, lutescenti-brunea, nitens ; 
albumen normale. 
Speciebus ab omnibus e subsectione Oblongiflorae distat imprimis petalis ovato- 
rotundatis. Inter reliquas forte juxta A. echinatam, Dun., intercalanda, sed multis 
de signis ab iUa abhorret. [Guiana.] 

Xylopia grandifloea, A. St.-Hil. Fl. Bras. Merid. i. p. 40. 

Mab. Ad Serra da Chapada mens. Aug. fructificat. (N. 202.) [Bras. Or., 
Amazonia.] 

seco:nd series. — botany, vol. iv. 2 s 



S06 ME. SPENCER LE M. MOOEE — PHANEROGAMIC BOTANY 

MENISPEEMACEiE. 

CissAMPELOS Paheira, Linn. Sp. PL ed. I. p. 1031. 

Sab. Ad Santa Cruz floret mens. Oct. (N. 175.) (In tropicis late disseminata.) 

CissAMPELOS TEOP^OLiFOLiA, DC Syst. 1. p. 532. 

Hah. Eeperi ad Santa Cruz mens. Nov. florentem. (N. 555.) [Peruvia.] 

CRUCIPER^. 

IsTASTrETiUM PUMiLUM, Cambess. in St.-Hil. El. Bras. Merid. ii. p. 119. 

Kah. Eloret et fructificat ad Santa Cruz mens. Nov. (N. 479.) [Bras. Or., Goyaz.] 

CAPPAHIDE^. 

Capparis Cynophallophora, Linn. Sp. PL ed. II. p. 721, var. microphylla. 

Hah. Inter Corumba et Ladario reperi mens. Jan. fructificantem. (N. 1045.) 
[Amer. Trop. Austr., Ind. Occ] 

Crat^va Tapia, Linn. Sp. PL ed. I. p. 444. 

Hah, Ad Corumba floret et fructificat mens. Jan. (N. 949.) [Amer. Trop., Ind. Occ] 

YIOLACE^. 

CoRYNOSTYLis PUBESCENS (sp. nov.) ; caule folioso, subtereti, in longitudinem rimoso, 
crebre lenticellifero, pubescente, demum glabro ; foliis sat parvis (maxime insequa- 
libus), breviter petiolatis, ovato- vel obovato-lanceolatis, obtusis vel obtusissimis vel 
8equaliter insequaliterve retusis, margine leviter undulatis seu serrulatis, pag. sup. 
fere glabris, inf. pubescentibus vel puberulis ; pedunculis gracilibus, folia excedentibus 
vel subsequantibus, pubescentibus ; sepalis ingequalibus, oblanceolatis vel oblongo- 
ovatis, obtusis ; petali antici limbo ovato, obtuso, margine undulato ; tubo insuper 
amplo hinc deorsum sensim angustato, apice dilatato ; pet. lateralibus late oblongis, 
margine undulatis, pet. posticis quam lateralia multo brevioribus, obovatis, obtusis ; 
staminibus oblongo-linearibus, quam sepala longioribus, apiculo rotundato-reniformi 
coronatis ; ovario oblongo compresso ; stylo stamina suba^quante, basi filiformi, 
superne incrassato ; capsula non visa. 
Hah. In campis juxta ripas fl. Paraguay ad Caran dazinbo, inter Corumba etDourados, 
mens. Januar. simul floret et fructificat. (N. 910.) 

Suffrutex parvus, crebre ramosus, circa 1 met. alt. Eolia l'8-5-5 cm. long., l'0-30 cm. 
lat., coriaceo-membranacea, eximie nervoso-reticulata, pagina superiore nitentia; 
costse secundarise utrinque 5-6, aperte arcuatse, sub margine tandem conjunctae ; 
petioli ad 0-5 cm. long., pubescentes. Pedunculi usque ad 5*0 cm. long. Sepala 
0-3-0'35 cm. long., O'l-O'lS cm. lat., uninervia. Elores albi, odorati. Petali antici 
limbus 0-7 cm. lat., hujus calcar 2-5-2-7 cm. long., sub apice 0*4 cm. lat., parte 
angustiore 02 cm. lat., rectum vel parum curvatum. Petala lateralia I'O cm., antica 



or THE MATTO GEOSSO EXPEDITION, 1891-92. 307 

vix 06 cm. long. Staminimi lociili fere 0*5 cm., apiciilus 0*1 cm. long., hie 

dimidio superiore auranliacus. 
Corynostylidl hyhantho, Mart., arete affinis, diversa videtur cquidem indumento, foliis 
minoribus aliquanto diversiformibus, corolla tiibo infra apicem insigniter angustato, &c. 
[Amazonia, Guiana, Nov. Granat., Venezuela, Ind. Occ] 

loNiDiUM LACTEX'M (sp. nov.). Suffrutex ? vix spitbameus, dense flavescenti-villosus, 
caulee rliizomate tenui albido obscure annulatim constrieto, erecto, levitcr angulato, 
ima basi cataphyllis parvis flavescenti-bruneis, sursum gradatim in folia trans- 
euntibus instructo ; foliis vix sessilibus, oblongo-ovatis, basi parum angustatis, antice 
dentato-serratis, utrinque flavescenti-villosis ; pedunculis quani folia brevioribus, 
villosis ; sepalis semipinnatiiido-pcctinatis, villosis ; petalis dorso villosulis, intus 
glabris, petali antiei, calycem sub 3-plo exeedentis, laminfi transversim oblonga. ; 
staminum anticorum filamentis dorso sub antliera incrassatis ; glabris, antlieris 
brevissime appendiculatis; ovario villoso. 
Hab. In campis ad Serra da Chapada alt. circa 1800 ped. Fl. Aug. (N. 187.) 
Folia ad 1*2 cm. long., in sicco subtus pallidiora. Flores lactei. Sepala usque ad 
0'7 cm. long., anticum falcato-lanccolatum reliquis paullo majus. rctalum 
anticum, calycem sub 3-plo excedens, ungue concavo angulato, calyci aequilongo, 
basi leviter dilatato, lamina transversim o])longa, obtusa, nervosa, margine crispula 
et dorso villosula, vix 1*5 cm. lat. ; intermedia duo antico baud multo breviora, 
lineari-oblonga, dorso villosula; postica duo sepala subsequantia, lineari-oblonga, 
obtusa, 1-nervia, dorso villosula. Genitalia calyce breviora ; antherse ovatse, apiculo 
tenuiter membranaceo spathulato parvo coronatse, loculis paullo obliquis ; filamenta 
antheris longiora, complanata, glabra. Ovarium ovoideum, dense villosum : stylus 
sursum dilatatus, glaber ; stigma obliquum. 
I. Ipecacuanli(B proxima, distat vero ob folia minora, sepala breviora, petalum anticum 
minus et obtusum (baud aperte emarginatuni), petala intermedia lineari-oblonga baud 
ovate-suborbicularia), antberas apiculo membranaceo loculis multi)ties breviore coronatas, 
filamenta staminum anticorum glabra. I. villosissimum, A. St.-HiL, ob pedunculum folio 
suffulciente longiorem, I. Foayam, A. St.-HiL, et J. lanatmn, A. St.-HiL, ob sepala integer- 
rima faciliter nostra a stirpe dignoscere potes. [Amer. Trop.] 

lONiDiUM COMMUNE, A. St.-HiL PL Rem. p. 295. 

JSaJ). Crescit juxta Corumba, mens. Jan. fiorens. (No. 1048.) [Bras. Or.] 

loNiDiUM opposiTiroLiUM, Rocm. & Schult. Syst. Veg. v. p. 395. 

Hab. Crescit juxta fl. Paraguay inter Santa Cruz et Villa Maria, itaque inter Villa 
Maria et Corumba. (Nn. 822, 1023.) [Amer. Trop. Austr.] 

Alsodeia, sp. nov. ? aff. A. ovalifoli^e, Britton in Bull. Torr. Bot. Club, xvi. p. 18. 

Sab. Viget in ripa fl. Paraguay inter Santa Cruz et Tres Barras, ubi fructificat mens.. 
Oct. (N. 502.) 

2s2 



308 ME. SPEyCEB LE M. MOOEE — PHANEEOGAMIC BOTANY 

This has almost exactly the leaves of Alsodeia antifolia, except for the ahsence of 
soft puhescence on the hack and on the petiole. The capsule is longer and nearly 
glahrous, and, if the only capsule on the type-specimen at Kew be mature, firmer in 
consistence, besides having narrower, more sharply acuminated and prominently reticu- 
lated valves. I have not seen flowers. [Bolivia.] 

SArvAGESiA ERECTA, Linu. Sp. PI. ed. II. p. 203. 

Eab. Ad Santa Cruz floret mens. Sept. (N. 754.) [Amer. Trop., Afr. Trop., Ind. 
Arch.] 

BIXINE^. 

OocHLOSPEHMTJM INSIGNE, A. St.-Hil. PI. Us. sub tab. 57, var. ? An sp. diversa ? 
Sab. Elorebat et fructificabat, foliis orbum, juxta Cuyaba mens. Aug. (N. 29.) 
Dr. "Warburg, w^ho saw my specimens, considered it unsafe, in the absence of leaves, to 

absolutely identify them. [Amer. Trop.] 

POLYGALACE^. 

PoLYGALA KHODOPTERA, Mart, cx A. W. Bcnu. in Mart. El. Bras. xiii. pars iii. p. 11. 

Hah. Ad Santa Cruz flores dilute viridi-purpurascentes prsebet mens. Nov. (N. 669.) 
[Bras. Or., Goyaz, Amer. Centr.] 

PoLTGALA HIRSUTA, A. St.-Hil. El. Bras. Merid. ii. p. 45. 

Kah. Crescit ad Serra da Chapada. (No. 193.) [Bras. Or.] 

PoLTGALA HYGROPHILOIDES, (sp. nov.) ; caule quadrangular i, sursum sparse ramoso ; 
foliis brevissime petiolatis, infimis verticillatis, reliquis approximatis vel subdistanti- 
bus, lanceolatis, apice acute mucronatis; racemis terminalibus, elongatis, pedanculatis, 
cylindricis, densifloris; sepalis exterioribus oblongo-ovatis, apiculatis; alis corolla 
multo longioribus, late obovatis, obtusissimis; carina crista bipartita, segmentis iterum 
bifidis superne dorso instructa ; seminibus subcylindricis, liorum caruncula breviter 
biappendiculata. 
Kah. Matto Grosso {Leeson^ in Herb. Brit. Mus.). 
Planta spithamea, radice brevi sparsim ramoso instructa. Caulis in longitudinem 
striatus, obscure 4-alatus, superne puberulus. Eolia+2-0 cm. long., 0'4-0'7 cm. 
lat., glanduloso-punctata. Bacemi pedunculis 2'0 cm. long, attingcntibus insidentes, 
usque ad 60 cm. long., vix I'O cm. diam., sursum angustati, apice inconspicue 
comosi, deorsum cicatriculis florum dilapsorum muniti. Bractese lineari- 
lanceolatse, longe acuminatse, dorso obtuse carinatse ibidemque nigro-glandu- 
lossB. Sepala exteriora insequalia, carinata, sep. maximum 0*2 cm. long. ; alse 
0-33 cm. long., 0"27 cm. lat., dorso carinulatae et nigro-glandulosae. Petala oblongo- 
ovata, obtusa, vix usque ad medium tubo staminali adnata, carinam galeatam in 
fructu persistentem superantia. Antherae filamenta libera subaequantes. Stylus 
brevis, crassiusculus, ovario sequilongus. Stigma cucullatiim et crista elongate 



OF THE :\[ATTO GROSSO EXPEDITION, 1891-92. 309 

superne prseditum. Capsula cordato-suborbicularis, membranacoa, 0*2 cm. long. 
• Semina fusco-brimea, liispidula, O'lS cm. long., OOS cm. lat. ; ciiriincula appendices 

late oblongse, quam semen saltern triplo breviores. 
Polygala Timoutou, Aubl., P. timontoidl, Cliodat, et P. hygrophilcc, H. B. K., affinis, 
ab illis vero primo obtutu secernenda ob liabitiim ramosum et racemes elongates 
pedunculatos. Ceterum P. timoutoides sepala exteriora calva, carince cristam parvam, 
necnon seminum minorum camnculam longe appendiculatam ostendit, et P, Timoutou 
racemos conspicue comosos, alas acutas, et semina appendieibns seipsam saltern semiiP- 
quantibus instmcta. P. hygrophila, planta elatior, foliis angiistis, racemis ct floribus 
minoribus, carinse crista dentata, seminum caruncula longius appeudiculata gaudet. 
[Amer. Trop.] 

Polygala angulata, DC. Prod. i. p. 328. 

Hah. Reperi ad Serra da Chapada. (N. 188.) [Amer. Trop.] 

VOCHYSIACEiE. 

Callisthene FAscicrLATA, Mart. Nov. Gen. ct Sp. i. p. ] 26. 

llah. Ad Jangada flores suaveolentes precbet mens. Sept. ; mens. seq. floret ad Santa 
Cruz. (Xn. 260, 539.) [Bras. Or., Goyaz, Bolivia.] 

Callisthene, sp. nov., ex affinitate praecedentis, abs qua diversa videtur ob folia majora 
(usque ad 16*0 cm. long, et 7"0 cm. lat. interdum attingentes) per paria plcrumque 
4-8 ramulis florentibus insidentia. Plores non suppetebant. Pructus admodum 
minutissimi solummodo obvii, cylindrici, glabri, rugati, 0*5 cm. long., pedunculis 
crassiusculis pubescentibus usque I'O cm. long, fulti, et stylo paulisper persistente 
ssepe incurvo lis ipsis aequilongo coronati. 
-Er«6. Prutex altus, vel potius arbuscula ; crescit ad Santa Cruz. (N. 731.) [Bras. Or., 

Goyaz, Bolivia.] 

Qtjalea grandifloka, Mart. Nov. Geo., et Sp. i. p. 133. 

Hah. Crescit juxta Santa Cruz, mens. Nov. florens. Pautej^a incolarum. (N. 740.) 
The popular name, spelt as above, was given me by an intelligent inhabitant of Santa 
Cruz. Martins spells it Pflo terra, and this I suspect to be the correct rendering. The 
fruit is mashed up and ap^Dlied to sores at Santa Cruz. [Bras. Or., Goyaz, Amazonia.] 

Qualea pakviflora, Mart. Nov. Gen. et Sp. i. p. 135. 

B.ah. Ad Santa Cruz mens. Oct. florescit. (N. 186.) 

Mine is the " forma glaberrima '* — very different in its leaves from typical Q. parviflora, 
but apparently identical in its flowers. [Bras. Or.] 

Qtjalea pilosa, Warm, in Mart. PL Bras. xiii. pars ii. p. 45. 
Sal). Crescit ad Santa Cruz, Dec. florens. (N. 751.) 

VocHYSiA DiVERGENS, Pohl, PI. Bras. Ic. ii. p. 19 (ex icon, et descript.). 
Hab. Matte Grosso {Leeson, in Herb. Mus. Brit.). [Bras. Or., Goyaz.]. 



310 ME. SPENCEE LE M. MOOEE — PHANEEOGAMIC BOTANY 

Salvertia. convallariodoea, a. St.-Hil. in Ann. Mus. Paris, vi. (1820) p. 266. 

Hah. Viget passim prope Cuyaba, fructus gerens mens. Aug. (N. 28.) [Eras. Or., 
Goyaz, Amazonia.] 

CARYOPHTLLE^. 

PoLTCARPON APURENSE, H. B. K. Nov. Gen. et. Sp. vi. p. 33 [Fharnacetmi depressum, 
Linn.). 
Hah. Crescit ad Santa Cruz, ubi floret mens. Nov. (N. 601.) [In tropicis utriusque 
orbis late dispersa.] 

PORTULACACEiE. 

PoRTTJLACA OLERACEA, Linn. Sp. PI. ed. I. p. 445. 

EaJj. Reperi ad Santa Cruz. (N. QQL) [Latissime disseminata.] 

PoRTULACA PiLOSA, Linn. Sp. PL ed. I. p. 445. 
Hab. Viget ad Puerto Pacheco. (N. 1052.) 
I refer my specimen to Rohrbach's Lusus L, forma a. [Amer. Trop. et Subtrop.] 

Talinum crassifolium, Willd. Sp. PL ii. p. 864. 

Hah. Floret ad Corumba mens. Jan. (N. 1030.) [Amer. Trop. Austr., Ind. Occ.] 

HYPERICACE^. 

ViSMiA DECIPIENS, Cham, et Schleclit. in Linnsea, iii. p. 116. 

Sal). Reperi in cacumine montium Serra da Chapada. (N. 137.) [Bras. Or.] 

ViSMiA JAPURENSis, Reichardt in Mart. PI. Bras. xii. pars i. p. 209, ex icon, et descript. 
Hah. Ad Santa Cruz floret mens. Nov. (N. 609.) [Amazonia.] 

GUTTIPER^. 

Rheedia GrAcoPARY (sp. nov.). Glaber, caule erecto, ramoso, plus minus quadrangular!, 
in sicco manifeste canaliculato ; f oliis petiolatis, oblongo-ovatis, acutis, basi rotundatis 
vel paullo coarctatis, margine undulatis vel undulato-repandis, coriaceis ; pedunculi& 
quam petioli brevioribus, plerumque solitariis ; petalis ovatis, obtusis, reflexis ; toro 
stamineo inferne lato, nequaquam acuto ; staminibus circa 19, quam petala brevi- 
oribus; ovario depresse ovoideo, staminibus plerisque longiore, 1 -2-loculari ; fructu 
globoso 1-spermo. 
Hab. Frequens ad Santa Cruz et juxta ripas vicinas fl. Paraguay, ubi menss. Sept.-Nov. 
fiores et fructus simul gerit. Guacopary incolarum. (Nn. 332, 365, 467.) 
Frutex altus vel arbor parva, ad 9 met. alt. Rami crassi, nonnunquam fere teretes, 
cinerei; ramuli ssepe per paria supra axillares orti. Folia maxime ina)qualia, 
3-0-15-0 cm. long., medio 1-5-7-5 cm. lat., obscure insequilateralia, baud in sicco 
nitentia, nee subtus decoloria, nervi omnes utrinque eminentes; petioli 0-7- vix 
2-0 cm. long., pingues, basi ampliati, canaliculati, in sicco transvcrsim corrucjati. 



or THE MATTO GEOSSO EXPEDITION, 1S91-92. 311 

Pedunculi e nodis plus minus tumidis proTCiiientes, circa I'D cm. long, et 0-05 em. 

crass., in longitudinem striati. Flores 0'7 cm. diam., albi, inodorati. Pctala 0-5 cm. 

long., 0*3 cm. lat., minutissime pelliicido-punctata. FilamentadivcrsaDlongitudinis, 

longiora 025 cm., breviora nee ultra O'lo cm. long., inferne parum ampliata ; 

antherse didymse, minutgp ; torus stamineus circa 0'075 cm. alt. Ovarium 0*2 cm. 

long., sub stigmate eminente parum coarctatum. Fructus magnitudine nunc baccam 

Mihis GrossularUe, nunc drupam parvam Ptmni domestical a^quans, luteus, exocarpio 

tenui coriaceo nitido obtectus ; semen 1-5 cm. long., 07 cm. lat., conipressum, succo 

subacido grato circumfusum ct endocarpio membranaceo-carnoso albo inclusum. 

Videtur ad H. Gard^ierianam, Planch. & Triana, et R. Spruceanam, Engl., proxime 

accedere, abs qua, notas inter alias, foliis diversiforniibus necnon floribus majoribus 

stirpem nostram facile dignoscere potes. 

Of this plant, which appears in two forms, one Avith smaller leaves narrowed ])elow, 
the other having larger leaves rounded at the base, the fruit is highly estc^cmed at Santa 
Cruz, and its juice is frequently employed in place of vinegar ; the berries also are said 
to be a good bait for certain of the Paraguay fish. [Bras. Or., Amazonia, Peruvia.] 

Platunia ?, sp. nov. Arbor data, glaberrima, ramulis subteretibus, inter se 4*0-7'0 cm. 
distantibus, foliigeris ; foliis oblongis vel obovato-oblongis, obtusis vel retusis, 
basi obtusis, 15'0-18'0 cm. long., 5*0-7'0 cm. lat., vivis saturate viridibus, in sicco 
pallidis; costis secondariis distantibus circa 8, angulis latis insertis, fere rcctis, demum 
subito undulato-arcuatis, supra impressis, subtus parum eminentibus ; petiolis vix 
usque ad 1*0 cm. long, crassiusculis, longitrorsum rugatis, superne anguste canali- 
culatis; floribus ignotis; baccis indehiscentibus, pyriformibus, magnitudine Fici, 
aurantiacis, coriaceis, intus carnosis, 2-4-locularibus, usque ad 5*0 cm. long., 40 cm. 
diam. ; seminibus 2-4 in pulpo sapido jacentibus, oblongis, usque ad 2*0 cm. long., 
1-0 cm. diam. 
Sab. Crescit in ripa fl. Paraguay ad Santa Cruz et alibi in vicinia, ubi mens. Nov. 
fructificat. Bacury incolarum. (N. 462.) 

According to authors, Bacury is the name given to the fruit of Platonia insignis. Mart., 
and of Bheedia macrophylla, Planch. & Triana. The present plant is, hoAvever, quite 
different from either of these, the venation of its leaves being more like what we find in 
the genus Tovomita, although, from the nature of the seed, I conclude that it cannot be 
placed in the latter genus. From the fact of our Paraguayan peons knowing this fruit 
well, it is to be presumed that the tree has a considerable range to the southward. 

TERNSTR(EMIACE^. 

KiELMEYEEA (§ CoRYMBosiE) AMPLEXiCATJLis (sp. nov.). Suffrutcx, caule crccto, crassi- 
usculo, tereti, glabro ; foliis vix omnino sessilibus, oblongis vel oblanceolato-oblongis, 
obtusis vel obtusissimis, leviter amplexicaulibus, coriaceis; glabris ; paniculis corymbi- 
formibus, folia hand excedentibus, foliis hebetatis interjectis ; sepalis subaequalibus, 
ovato-lanceolatis, acutiusculis, dorso carinatis ; petalis o-vatis vel ovato-oblongis, obtusis, 



312 ME. SPENCEE LE M. MOOEE — PHANEEOGAMIC BOTANY 

vel obscure retusis, "breviter lateque unguiculatis, planis vel carinatis ; staminibus 
quam petala dnplo brevioribus ; antberis linearibus, obtusis, apice glandula miauta 
instructis; conuectivo latiusculo; ovario ovoideo trigono, in stylum crassimi staminibus 

subaequilong-iim subito desinente ; stigmate depresse pyramidali ; fructu 

Hah. Viget in campis ad Serra da Chapada, alt. circa 1800 ped. supra mare. M. Aug. 

(N. 183.) 

Crescit circa ad f met. alt. Caulis simplex, basi incrassatus, e rliizomate lignoso cortice 
squamoso obtecto, plenitudine pennam anserinam duplo excedente. Folia plerumque 
10'0-12*0 cm. long., 3'5-4*5 cm. lat., nervo medio subtus eminente, nervis laterali- 
bus impressis, copiose glandulosa; folia hebetata reliquis subconformia, 0'6-l*7 cm. 
long. Panicula? ad lO'O cm. long,, circa 5-0 cm. diam. Mores albi, odorati. 
Sepala 0*8 cm. long., 3'0-4"5 cm. lat., glabra vel medio dorso leviter rufo-puberuia. 
Petala fere omnino glabra, margine undulata, 1*7 cm. long. Ovarium 04 cm. long. ; 
stylus ovario suba^quilongus. 
A Kielmeyerd corymhosd, Mart., cujus vix est varietas, differt prsesertim foliis 

amplexicaulibus et petalis majoribus glabris. [Bras. Or., Goyaz.] 



MALVACE^. 

SiDA COEDIFOLIA ?, Linn. Sp. PL ed. I. p. 684. 

Sab. Crescit ad Santa Cruz, ubi Tcmiarindo do CamiJO nuncupatur. 

The leaves of this plant, boiled, are used as an emollient. [In regionibus tropicis 
et subtropicis vulgatissima.] 

WisSADTJLA DECOEA (sp. nov.). Tcnuiter lanato-tomentosa, caule copiose ramoso, sub- 
tereti ; foliis petiolatis ovatis, nonnunquam vero ovato-lanceolatis, obtusis, basi alte 
cordatis, insequaliter vel subsequaliter crenato-dentatis, 6-7-nerviis, stipulis setaceis 
petiolo multo brevioribus ; floribus in paniculas subbreves digestis ; pedicellis pro 
genere brevibus, calycis tomentosi lobis deltoideis, acutatis, subsequalibus ; petalis 
calycem 3-plo excedentibus ; androecio quam calyx longiore ; ovario 3-mero, 3-locu- 
lari, glabro ; stylis 3 basi ipsa connatis ; ovulis 2 vel 3 pendulis : si 2, ambobus juxta 
apicem loculi affixis ; si 3, duobus apicalibus, tertio in medio loculo posito. 
Hah. Abundat in campis juxta fl. Paraguay ad Puerto Paclieco. PI. Peb. (N. 1061.) 

Suffrutex ? parvus, erectus, decorus. Eamuli tenues, vix 0"2 cm. crass. Poliorum lamina 
l'5-3'5 cm. long., 1-0-20 cm. lat., subtus pallida ; petioli filiformes, I'O-ultra 25 cm. 
long., patuli; stipula? 0*2 cm. long., deciduse. Paniculge circa 10"0 cm. long., 
subpyramidales, multiflorge, graciles; pedicelli ad 2-0 cm. (plerumque vero circa 
0-5-1-0 cm.) long., filiformes, tomentosi ; bractese stipulis subsimiles, deciduae. 
Plores lutei, lo cm. diam. Calycis tubus 0*3 cm. long. ; lobi 0'14-0*2 cm. lat., vix 
0*2 cm. long. Petala obovata, obtusissima, delicatula, basi leviter pilosa, 0*8 cm. 
long. Andrcecium 0*4 cm., tubo circa 0*15 cm. long. Ovarium 0*1 cm. long., 
puberulum ; styli filiformes, androecio longiores, stigmatibus capitatis, circa 0-025 
cm. diam. Capsula deest. 



OF THE MATTO GROSSO EXPEDITIOX, 1891-02. 313^ 

Planta cum Wissaduld gymnanthemd, K. Schiim,, comparanda, specie quiB ovario 
trimero saepe gaudet, ab ilia vero optime abliorrot liabitu et foliis omnino disparibus, 
calyce minora, ovario nunquam tetramero, etc. 

This is a very elegant little plant, which, in the absence of capsules, I regret niv 
inability to assist in introducing into cultivation, 

Schumann (Mart. Fl. Bras. xii. pars iii. p. 438) correctly states that the ovules of this 
genus are all pendulous, and not, as say the authors of the * Genera Plantarura,' some pen- 
dulous and some ascending. With respect to the constant trimerous symmetry of the 
carpel, which is abnormal for the genus, I may here add that the carpellary symmetry of 
W, periplocifolia, Presl, said by Schumann to be pentamerous, is by no means so always, 
as on several occasions I have had before me 3- and 1-lobed carpels respectively with 
three or four styles. [Argentina.] 

Abutilon crispum, Sweet, Hort. Brit. ed. I. p. 53. 

Hab. Crescit ad Santa Cruz, mens. Sept. florens. (N. 296.) 

Mine is a monospermous form, the ovules being always solitary in the cells, and the 
seeds of course solitary. Moreover, the stems, petioles, and fruits are pubescent merely, 
never hirsute, as is almost always the case with A. crispum. The plant is certainly not a 
good Abutilon, and perhaps it would be better to keep up the genus Bclcerc, Shuttlew., 
for the reception of monospermous Malvaceae with Abutilon-like unappendaged seeds. 
That genus would then comprise, besides the present plant, one from Brazil recently 
described by Schumann under the name of A. mouospennum, and several others with 
carpels constantly 1-seeded. The attachment of the ovule, too, to the septum, instead of 
to the bottom or top of the cell, is a fact fuU of suggestion. This is a question, however, 
on which I am not disposed to give a definite opinion. [Amer. Trop. et Subtrop., 
Ind. Or.] 

Sph^ralcea miniata, Spach, Hist. Nat. Veg. iii. p. 352. 

, var. LEIOCARPA, S. Moore ; foliis ovato-oblongis, crenatis vel breviter lobulatis, supra 

minute pubescentibus, subtus molliter tomentosis ; inflorescentiis sessilibus vel 
breviter pedunculatis, floribus aggregatis ; carpidiis intus basi levibus vel apice 
obscure reticulatis, saepe 3-spermis. 
Sab. Matto Grosso (Leeson, in Herb. Mus. Brit.). [Argentina.] 

Pavonia speciosa, H. B. K. Nov. Gen. et Sp. v. p. 231. 

, var. POLTMORPHA, Garcke. 

Hab. Reperi ad Jangada, mens. Sept. florentem. (N. 281.) [Amer. Trop., Ind. Occ] 

Pavonia Morongii, S. Moore ex Morong & Britton in Ann. N. York Acad. Nat. Sc. vii. 
p. 58 ; caule erecto, ramulis subpatulis, ramoso, piloso-tomentoso ; foliis par\'is, petio- 
latis ovatis, obtusis, basi rotundatis necnon aliquanto cordatis, margine impariter 
ffrosse crenato-serratis, 7-nerviis, tenuiter tomentosis, subtus albidis ; stipulis fugaceis, 
subulato-setaceis, petiolo brevioribus ; floribus in axillis fol. superiorum vel ad apicem 
ramulorum solitariis, longius pedunculatis ; involucri phyllis 6-9, tomentosi-pilosis, 

SECOND SERIES. — BOTANY, VOL. IV. 2 T 



314 ME. SPEXCEE LE M. MOOEE— PHANEROGAMIC BOTANY 

calyce circa duplo longioribiis ; calycis tubo ampliato, basi 10 superne 15-nervato, 
lobis ovato-lanceolatis tubo subsequilongis ; fioribus circa 1*2 cm. diam. ; carpidiis 
calvis, maxime depressis, ecarinatis, facile in valviilas duas dehiscentibus. 
Hab. Crescit prope Corumba sine numero ; fl. Jan. 

Suifrutex parvus, vix ^ met. alt. Eadices albi, sparsim iibrosi. Caulis demum fere 
omnino glaber ; pallide bruneus, juxta basin crebrc longitudinaliter rimosus ; 
ramuli graciles, teretes. Polia l* 2-2-0 cm. long., medio l'O-2-O cm. lat. ; stipulse 
puberulge 035 cm. long; petioli ad 2*2 cm. long, (plerique vero breviores), graciles, 
pilosi-tomentosi. Pedunculi ad 3-5 cm. long, (plerumque circa 3*0 cm.), graciles, 
supra medium articulati, pilosi-tomentosi. Involucri phylla circa 1*0 cm. long., 
filiformia. Calycis piloso-tomentosi tubus 0*3 cm. long. ; lobi 01 cm. long., 3-nervati. 
Plores lutei. Carpidia 0'4 cm. long, et 0'4< cm. lat., venosa (nervo medio parum 
eminente), in sicco virentia. Semina obconica, vix 0"3 cm. diam., glabra. 
Affinis Favonice JBalcmscB, Guerke, et P. JHeronymi, Guerke, sed multis notis ab 

ambabus dispar. 

The easy splitting of the carpidia, seen also in the case of P. BalmiscB, is curious, and 

might perhaps justify the creation of a new genus. The greatly depressed form of the 

same organs is also unlike what we usually find among the Pavonias. [Paraguaria.] 

Pavonia Mutisii, H. B. K. Nov. Gen. et Sp. v. p. 283. 

, var. HEXAPHYLLA, S. Moorc. A typo discrepans vestitu pilorum patentium, habitu 

gracili, pedunculis tenuibus, involucri phyllis plerumque 6, raro 7, angustis (0"05 cm. 
lat.), sed vix, me judice, species propria. 
Hah. Crescit ad ripam fl. Paraguay, prope Barra do Sao Lorenzo. (N. 10.) 
Hujus varietatis exstant formse duse, foliis alterius subtus cancscenti-tomentosis, 
alterius canescenti-pubescentibus, dein pubescentibus. Hsec involucri phyllis brevibus 
calycem vix excedentibus gaudet, illius involucri phylla contra calyce duplo longiora. 
[Bras. Or., Colombia.] 

Pavonia Eosa-campesteis, A. Juss. in A. St.-Hil. Plor. Bras. Merid. i. p. 236. 

Hal). Crescit in cacumine montium Serra da Chapada, mens. Aug. florens. (N. 181.) 
[Bras. Or.] 

Pavonia optjlipolia (sp. nov.) ; caule erecto, subtereti, stellatim tomentoso, mox 
appresse pubescente ; foliis petiolatis, late cordatis, 3-5-lobatis (plerumque 3-lobatis), 
impariter grosse crenato-serratis, 7-9-nerviis, sparsim stellatim pubescentibus et ad 
nervos tomentosis; stipulis laiiceolatis ovato-lanceolatisve, quam petioli brevioribus; 
fioribus in axillis fol. superiorum solitariis; pedunculis quam petioli longioribus; 
involucri phyllis 13-14, calyci subsequilongis ; calycis lobis lanceolatis, tubo 
longioribus ; corolh\ vix 20 cm. diam. ; tubo stamineo 0*6 cm. long. ; stylis obscure 
puberulis ; carpidiis calvis, medio dorso leviter carinatis. 
Hab. Inveni mens. Aug. ad ripas fl. Cuyaba vigentem, loco longinquo ab urbe Cuyaba 

iter duorum dierum. (N. 3.) 



OF THE MATTO GROSSO EXPEDITIOX, 1891-92. 315 

Siiffrutex incertie altitudinis, erectus. Internodia 1 •0-2-7 cm. long. Folia 2'0-5'0 cm. 

long., medio 2'0-4"5 cm. lat., subtus decoloria ; petioli l'2-20 cm. long., patuli, 

tomentosi; stij)ul9B 0-4 cm. long., scarioste, pubescentes. Pedunculi 15-2-5 cm. 

long., tomentosi. Involucri pliylla anguste linearia, tomcntosa, usque ad 11 cm. 

long. Calycis tomentosi tubus amplus, 10-nervosus, 03 cm. alt. ; lol)i vix OS cm. 

long., 3-nervosi, recte vel oblique acuti. Petala lutea, parum obliqua, obovata, 

obtusissima, breviter lateque unguiculata, circa 20 cm. long., medio I'O cm. lat. 

Andrfficium circa 15 cm. long. Styli andrcccium paullo excedentes. Carpidia 

0"5-0"6 cm. long., medio dorso 035 cm. lat., lateribus eximie venosis fere cristatis, 

apice obtusa. 

Affinitatem prsebet cum Pavonid humifusd, A. Juss., quae liabitu dispari, foliis 

majoribus longius petiolatis baud impariter crenato-serratis, internodiis longioribus, 

stipulis subulato-filiformibus, involucri phyllis 14-18 inter alia propria gaudct. 

Pavonia opulifolia, var. major, S. Moore. A typo divergcns babitu robustiore, statural 
altiore ? (ultrametrali), floribus majoribus, involucri ^^byllis calyce i'crc duplo longi- 
oribus (vix ad 2*0 cm. long.) hirsutis, necnon carpidiis paullo majoribus (0"G-07 
cm. long.) apice acutis, in sicco fuscioribus. 

Hah. Crescit in campis ad ripas fl. Paraguay juxta Corumba, ubi mens. Jan. flores et 
fructus praebet. (N. 913.) 

At first sight this j)lant and the type seem very unlike, but close examination reveals 
so many points of agreement that I refrain from making two species of them. The 
specimens which have served for the description of the type are small pieces hastily 
gathered, on our way up to Cuyaba, at one of the log-cuttmg settlements ; hence they 
probably do not do full justice to the species — an additional reason, this, for the course 
here taken. The diiference in time of flowering of type and variety is noteworthy. 

In the description of the length of the staminal tube (in species of Pavonia) I have 
measured only the distance between its bottom and the point whence the first batch of 
stamens is given off. [Bras. Or.] 

Pavonia velutina, A. Juss. in A. St.-Hil. PL Bras. Merid. i. p. 233. 

Mab. Pructificat ad Serra da Chapada mens. Aug. (N. 140.) [Amer. Austr. Trop.] 

CiENFUEGOSiA PHLOMiDiEOLiA, Garcke, in Bonplandia, viii. p. 148. 

Mab. Ploret ad Jangada mens. Sept., itaque ad Santa Cruz mens. seq. (Nn. 270, 452.) 
[Amer. Austr. Trop.] 

CiENPTiEGOSiA SULPHUREA, Grarckc, in Bonplandia, viii. p. 148. 
ffab. Eeperi ad Puerto Pacheco mens. Peb. florentem. (N. 1056.) 
The first recorded tropical habitat for this plant. [Uruguaria, Argentina, Paraguaria.] 

BoMBAX ?, sp., forsan ex affinitate JB. parviflorce. Mart, et Zucc. Arbor 6-metralis, capsulis 
oblongis lana fulvesceuti repletis, valvis 8*0 cm. long., 1-5-2-0 cm. lat., nitidulis, extus 

2t2 



316 MR. SPENCEE LE M. MOOEE— PHANEROGAMIC BOTANY 

ochraceo-bruneis, intus cinereis ; seminibus, magnitudine pisi, testa brunea obtectis. 

Nee floras nee folia suppetebant. 
Hab. Ad Cuyaba mens. Aug. fructificat. (N. 41.) 
The wool is used for stuffing purposes at Cuyaba. 

STEUCULIACE^. 

Helicteres guazum^folia, H. B. K. Nov. Gen. et Sp. v. p. 304. 

Sah. Crescit ad Santa Cruz, necnon juxta Corumba. (Nn. 383*, 904, 1002.) [Amer. 
Trop.] 

Helicteees cohylifolia, Nees & Mart, in Nov. Act. Acad. Nat. Cur. xii. p. 44. 
Hab. Ad Santa Cruz flores prsebet menss. Sept.-Nov. (N. 294.) [Bras. Or.] 

Helicteres chapadensis (sp. nov.) ; caule erecto, tereti, sparsim foliato, molliter 

lanato-tomentoso, demum fere glabro ; foliis parvis, petiolatis, ovatis, obtusis vel 

obtusissimis, margine impariter serratis, molliter ac tenuiter lanato-tomentosis ; 

stipulis subulato-filiformibus, petiolo fol. evolutorum brevioribus, firmis, tomentosis, 

deciduis ; dicliasiis 2-3-floris ad apicem ramulorum brevium digestis ; prophyllis 

filiformibus, tomentosis, caducis ; floribus horizontalibus ; calycis elongati, anguste 

tubulosi, basi sensim coarctati, lanato-tomentosi lobis insequalibus lanceolatis vel 

ovato-lanceolatis, acuminatis, tubo 3-plo brevioribus ; petalis calycem circa J exce- 

dentibus, anguste spathulatis, obtusis, longius unguiculatis, supra unguem dentibus 

2 pauUo curvatis auctis vel omnino calvis ; gynophoro calycem ultra 3-plo excedente, 

curvato, filiformi, glabro ; tubo stamineo vix evoluto ; staminodiis spathulatis, 

acutis ; ovario quam calyx multo breviore, oblongo, leviter contorto ; stylis connatis, 

tortis ; fructu ignoto. 

Hah. In convalle ad Serra da Chapada reperi mens. Aug. (N. 76.) 

Suffrutex erectus, 1 met. alt., inferne aphyllus, sparsim ramosus. Bamuli breves (nee 

ultra 1"0 cm. loug.), ex axillis fol. dilapsorum soli floriferi. Eoliorum lamina ad 

3'0 cm. long., medio ad 3*2 cm. lat., subtus eximie reticulato-venosa et 5-nervia; 

petiolus ad 1'3 cm. long., pinguis, lamina sub ipsa incrassatus, lanato-tomentosus ; 

stipulse circa 0-5 cm. long. Pedicelli 0*5-0'7 cm. long., 0"075-0-l cm. crass., 

tomentosi. Prophylla ad O'S cm. long., plerumque incurva. Calycis tubus l-3-l*6 

cm. long., ima basi tantummodo 02 cm. lat., superne usque 0*4 cm. amplificatus, 

hand nervosus; lobi circa 0*5 cm. long. Petala 2-2-2*4 cm. long., coccinea, eorum 

unguis 0-7 cm. long. Gynophorum 6-5-7-0 cm. long., superne attenuatum. Stamina 

10 ; antherse 0*13 cm. long. ; staminodia antheris subsequilonga. Ovarium vix 

02 cm. long. 

Cum H. molli, K. Schum., comparanda, abs qua nuUo negotio dignoscenda, imprimis 

foliis minoribus, subtus eminenter nervosis, hand argute serratis, calycis tubo coarctato, 

etc. [Bras. Or., Goyaz.] 

Helicteres brevispika, A. St.-Hil. PL Bras. Merid. i. p. 274. 

JETab. Ploret ad Santa Cruz mens. Oct. (N. 650.) [Bras. Or., Nov. Granat.] 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 317 

Helicteres orthotheca (sp. nov.) ; Ccaule erecto, tereti, tomentoso, demum fere glabro, 
cortice subcinereo, crcbro lenticellifero, anguste elliptico rimoso ol)tecto ; ramulis 
tomentosisjcopiose foliatis; foliis potiolatis, ovatis, aciitis vol obtusis, iinparitcrcronato- 
serratis, supra stellatim pubescent! bus, subtus albide toinentosis ; stipub's subulato- 
filiformibus, quam petioli brevioribus, concinnis, pedunculatis,4-7-floris, nonnunquani 
apicem versus ramulorum subcongestis j prophyllis caducissimis; floribus horizonta- 
libus, calyce amplo tubuloso incequaliter 5- (an semper ?) lobo tomentoso, lobis tubo 
3-plo brevioribus ; petalis calycem circa ^ excedentibus, spatbulato-panduriformil)us, 
apice bilobis, ungue quam lamina duplo breviore ; gynophoro calycom 4-plo exce- 
dente, curvato, in sicco plerumque subligulato, glabro ; tubo stamineo evoluto ; stami- 
nodiis lineari-spathulatis ; ovario quam calyx multoties breviore, oblongo, recto ; 
sty lis rectis; fructu elongate, recto, 5-3 era. long. 
Hab. Frequens in fruticetis ad Santa Cruz, ubi fl. et fruct. gerit menss. Nov. et Dec. 

(N. 556.) 

Frutex circa ad 3 met. alt. Caulis circa 0*6 cm. diam. Foliorum lamina 3'0-8*0 cm. 
long., 2"0-7'5 cm. lat., 7-8-neryia, cliartacea ; petioli 1-0-1-5 cm. long., tomentosi. 
Stipulse circa 0'8 cm. long., tomentosse. Pedicelli 0'2-0'7 cm. long., tomentosi. 
Calycis utrinquc tomentosi tubus vix 1'5 cm. long., medio 0*8 cm. lat., eminenter 
10-nervatus ; lobi 0"5 cm. long., late ovati, obtusi vel obtusiusculi, 0'4-0"G cm. long. 
Petala 2'5 cm. long., juxta apicem 0"7 cm. lat., rubra. Gynophorum vix 8'0 cm. 
long., vix O'l cm. lat., sub fructu 0*35 cm. lat. Tubus stamineus 0"3 cm. lat., et circa 
0'3 cm. long., late cyathiformis, glaber, eminenter nervosus. Stamina 10. Stami- 
nodia O'i cm. long., l-nervosa. Ovarium vix 0*3 cm. long., tenuiter tomentosum. 
Eructus oblongi; valvular apice sat subito coarctatae, dorso tomento tenui gaudentes, 
eminenter nervosae, leviter carinatae. 
Species ob fructum magnum, elongatum, rectum nuUi nisi S. Rushyi, Britton [Balcmsa, 

n. 1602), affinis, cujus folia attamen disparia longius petiolata utrinque tomentosa, 

gynophorum longius, fructus brevior (4'5-5 cm. long.), etc. [Bolivia, Paraguaria.] 

Melochia GRAMiNiFOLiA, A. St.-Hil. Flor. Bras. Merid. i. p. 160. 

Hab. Crescit ad Jangada, itaque ad Santa Cruz. (Nn. 258, 311.) [Bras. Or., Goyaz, 
Nov. Granat., Guiana.] 

Melochia (§ Kiedlea) corumbensis (sp. nov.). Suffruticosa, moUiter pubescenti-tomen- 
tosa, caule stricto subtereti; foliis petiolatis, oblongo-ovatis, obtusis, basi rotimdatis vel 
obtusis ; irregulariter dentatis vel dentato-crenatis ; inflorescentia petiole breviore 
(nonnunquam vero subaequali vel longiore), cymosa, pluriflora ; prophyllis linearibus, 
acutis; calycis cupuliformis lobis subdistantibus, deltoideo-acuminatis; petalis obovato- 
oblongis, obtusis, basi breviter unguiculatis, glabris ; tubo stamineo puberulo (floris 
brevistyli petalis bre^dore) ; ovario sessili. 
Sab. Yiget in locis planis juxta Corumba, mens. Jan. tl. (N. 905.) Eandem 

plantam in Matto Grosso legit Leeson, in speciminibus in Herb. Brit. Mus. servatis. 

Caulis 3*0-4-5 cm. crass., arete in longitudinem sulcatulus. Stipulse decidual, linear!- 



318 ME. SPEIsTEE LE M. MOOEE— PHANEEOGAMIC BOTANY 

subulatsc, acuminatae, pilosi-tomentosse, circa O-Ji cm. long. Petioli l*2-3'2 cm. 

long., sub lamina paiiUo ampliati ibidemque nonnunquam arctissime bialati, piloso 

tomentosi; lamina 4!'0-7"0 cm. long., basi 5-8-nervis. Prophylla piloso-tomentosa, 

02 cm. long. Cymse ad 2-0 cm. diam. Calycis lobi 0*2 cm., tubus vix 0'3 cm. long. 

Petala 0*7 cm. long., pur2)urea. Filamenta fl. brevistylium gynsecio duplo longiora. 

Capsula hirsuto-tomentosa, oblate spbseroidea, pentagona, vix 0*5 cm. diam. Semina 

bilateraliter compressa, levia, nigra, 0*2 cm. long. 
IlelocMce cinerascenti, A. St.-Hil. et Naud., arete affinis, sed indnmento, cymis pluri- 
floris, floribus majoribus, calycis lobis deltoideo-acuminatis nee subulate -filiformibus, 
petalis longioribus, basi ipsa (nee abrupte mediocriter) unguieulatis, et staminum tubo 
puberulo vix conspecifiea. [Amazonia.] 

Waltheria viscosissima, a. St.-Hil. El. Bras. Merid. i. p. 150. 

Sah. Crescit prope Barra da S{\o Lorenzo. (N. 9.) [Bras. Or., Goyaz, Guiana, 
Nov. Granat.] 

Guaztjma tjlmifolia, Lam. Encye. iii. p. 52. 

Sab. Ad Santa Cruz floret mens. ISTov. (N. 578.) [Amer. Trop.] 

Byttxeeia campestris (sp. nov.). Arbusciila debilis, spinosa, raraosissima; ramulis 
patidis, pubescentibus, mox glabris et cinereis, internodiis brevibus, ssepe curvatis ; 
foliis parvis, coriaceo-membranaeeis, breviter petiolatis, oblongis, obtusis, basi parum 
rotundatis, integerrimis, supra glabris, subtus in axillis nervorum prsecipuorum 
barbatis, nervo medio subtus eminente ibidemque juxta basin glandula oblonga vel 
oblongo-ovali notato ; umbellis plerumque solitariis, 2-6-floris (rarissime 1-floris) ; 
pedunculis brevibus vel longiusculis ; calyce pubescente; petaloium ungue late lineari; 
alis exiguis, patentibus, subdeltoideis, glabris ; ligula filiformi, tenui, villosula : ovario 
ovoideo, tuberculato; stigmate obscurissime 5-lobulato. 
Rah. Crescit in campis ad ripas fl. Paraguay ad Carandajinho, inter Corumba et 

Dourados. (N, 1015.) 

Erutex ascendens, spinis minutis disjectis obtectus. Eoliorum lamina ad 4*7 cm. long., 
1 cm. lat. ; costa3 secundaria? utrinque 6-8, erecto-ascendentes, rarius leviter forni- 
catse, ipso sub margine subito arcuatim-conjunctse ; petioli vix ad 0*4 cm. long., 
puberuli. Stipulse minutse, subulatap, caducissimse. Pedunculi ad 0*8 cm. long., 
filiformes ; pedicelli 0*5 cm. vel ultra long., capillares, glabri. Calycis lobi ovato- 
lanceolati, acuminati, 035 cm., long., nervosi. Petala 0*6 cm. long. ; unguis 0"15 
cm. long. Tubus stamineus O'l cm. long., prima quarta parte divisa; antherse 
0-05 cm. long., 0"16 cm. lat., approximatae. Pistillum 01 cm. long., ovario duplo 
longius. 
B. Jilipedi, Mart., proxima, sed habitu ramosissimo, ramulis pubescentibus, spinis 

minntis, foliis diversiformibus, brevius petiolatis, nequaquam obliquis, basi glanduliferis, 

necnon flore dispari inter capita alia speciem propriam rite dijudicari. [Bras. Or., 

Paraguaria.] 



OF THE MATTO GROSSO EXPEDITION, 1891-92. 319 

Byttneria Leesoni (sp. nov.). SuflFruticosa, ramosa, inermis; ramulis 5-angiilatis vel 
5-quetris, glabris; foliis superioribiis solummodo visis,parvis, lincaribiis, brevipetiolatis, 
integerrimis, glabris, laevibus; inflorcscciitui ut in B. ntmosisnimd, Polil; calycis 
laciniis ovato-lanceolatis acuminatis, integris ; pctalorum cucullo oboYato ; alis pilosis ; 
ligula calycem bene superante, pilosa ; staminodiis tridentatis ; ovario globoso, tuber- 
culato, basi incrassato. 
Hcib. Matto Grosso {Lceson, in Herb. Erit. Mus,). 

Kamuli rigidi, usque 0"3 cm. diam., in sicco pallidc brunei. Folia rigida, modo 1*5 cm. 
long., vix O'l cm. lat., petiolis lamina? aequilatis vix ultra 01 cm. long, fulta, basi 
glandulosa. Pedunculi et pedicelli graciles, usque 08 cm. long., hi soDj^e sub flore 
incrassati. Calyx totus 0'5 cm., tubus 0"15 cm. long. Petalorum unguis vix 
0*1 cm. long. ; cucullus circa O'l cm. lat. ; ligula 0*7 cm. long., filiformis, basi 
ampliata. Tubus stamineus 0*13 cm. alt. Staminodia medio dorso dente minute 
munita. Pistillum totum 0*15 cm. long. Capsula 00 cm. long., 0'7 cm. lat., 
depressa, globosa, alba, tuberculis acutis, erectis, apice curvatis, fuscis, densiuscule 
obtecta. Semina 0*4^ cm. long., 0'2 cm. lat., pallide grisea, verrucis parvis nigris 
instructa. 
Ab affini B. ramosissimd, Pobl, abliorret foliis alienis, petalorum ligulis multo lon- 

gioribus et una cum alis pilosis, ovario majore basi incrassato, necnon caj^sulis aliquatcnus 

diversis. [Goyaz.] 

Byttneria muricata (sp. nov.). Fruticosa, ramosa, ramis et ramulis spinis abbreviatis 
parum recurvis sparsim vel sat abundc mimitis, his primo minute tomentosis, deinde 
puberulis vel glabris ; foliis omnibus longiuscule petiolatis, late cordatis, nunc obtusis, 
nunc obtusissimis, membranaceo-coriaceis, palmatim 5-7-nervibus, utrinque scabridis ; 
inflorescentia axillari vel terminal!; floribus ignotis; capsula globosa, tuberculis 
pyramidalibus, basi ampliatis, apice spinuloso-acuminatis dense muricata. 
Hab. Crescit ad ripas fl. Paraguay inter Santa Cruz et Diamantino, ubi mens. Oct. 

fructificat. (N. 636.) 

Rami teretes, arrecti, juveniles minute tomentosi, mox puberuli, dein scabriusculi et 
cortice purpureo-cinerascenti crebre longitrorsum rimoso obducti; rami usque 
0-5 cm., ramuli 0-1-0-2 cm. diam. Foliorum lamina 7'0-10-0 cm. long., 4-5-11-0 cm. 
lat. (mediocris 8-5x70 cm.), inermis, margine impariter crenulata, raro obscuris- 
sime lobulata, utrinque i^ilis stellatis depressis scabrida, subtus decolor; costse 
secundariae (basalibus exemptis) 4-5, patulo-ascendentes, usque ad marginem rectse 
vel fere rectse, rarius leviter arcuatse, costas rectas tertii ordinis abundanter emit- 
tentes, costse omnes supra planse, vix impressa?, subtus eminentes, et basi glandu- 
Uferse vel nonnullae glandula carentes, raro omnes eglanduliferae ; petioli 2-0-4-0 cm. 
long., minute tomenteUi, deinde pubescentes vel puberuli, subquadrangulares, 
breviter canaliculati, nunc aculeis paucis (1-2) parvis recurvis onusti, nunc inermes. 
Pedunculus sub fructu incrassatus, 0-5-0-8 cm. long., 0-15 cm. lat. Pedicellus sub 
fructu pedunculo ^uboequilongus et subsimilis. Capsula 1-5 cm. long., 1'3 cm. 
diam., scabrida, tubercula 0-15-0-2 cm. long., basi 5-6-angularis. Semina 0-7 cm. 



320 ME. SPENCER LE M. MOOEE — PHANEROGAMIC BOTANY 

long., 0*3 cm. lat., subtrigona, apice breviter spinuloso-acuminata, obsolete nigata^ 

brunea ; bilum albidum ; chalaza purjmrea. 
Planta forte cum Byttnerid catalpcefolid, Jacq., comparanda, specie inermi, cujus folia 
nequaquam scabrida, pedicelli et capsulse omiiino dispares, etc. [Bras. Or., Amazonia^ 
Peruvia, Venezuela.] 

Byttneeia charagmocaupa (sp. nov.). Suffruticosa, sparsim ramosa, inermis; ramis 
crebre foliatis, leyiter undulatis, subteretibus, pubescentibus, dcinde glabris ; f oliis 
omnibus breviter et patule petiolatis, oblongis vel aDguste oblongo-ovatis, obtusis, 
basi leviter coarctatis, subrotundatis, interdum levissime cordatis, coriaceo-membra- 
naceis, mox, axillis nervorum pag. inf. exemptis, glabris vel fere glabris ; costis 
secundariis 7-10 approximatis distantibusve, angulis latis insertis, rectis vel levissime 
arcuatis, juxta marginem subito fornicatim-conjunctis, subtus admodum, supra leviter 
eminentibus ; inflorescentia abbreviata, ex umbella solitaria brevipedunculata 4i-flora 
(an semper ?) composita, floribus parvis, calycis pubescentis lobis triangulari-lanceo- 
latis, breviter acuminatis ; petalorum cucullo late lineari, superne subito amplificato ; 
ligula e basi lata lineari-lanceolata, sursum acuminata, puberula ; alis parvis deltoideis 
incurvis; staminodiis breviter 3-lobulatis ; ovaxio globoso, albido-tomentoso, mox plane 
tuberculato ; stigmate obscure 5-lobulato ; capsula subglobosa, tuberculis subcylin- 
dricis admodum elevatis, apice clavellatis, densissime obsita. 
Hah, Scliedula prsetermissa, locum natalem indicare nequeo, sed verisimiliter crescit 

ad Santa Cruz vel alibi in vicinia. 

Bamuli 0'2-0'3 cm. diam., grisei, rimis anguste ellipticis abundanter insculpti, intervallis 
plerumque 1*0-1*5 cm. long, foliigeri. Polia pleraque 7'5-10'0 cm. long., et 3*5- 
5'0 cm. lat., margine tenuiter cartilaginea ibique obsolete undata ; costa media 
dorso basi glandulifera ; petioli plerumque 0'5-0"7 cm. long., patuli, saepe decurvi, 
supra breviter canaliculati, pubescentes vel puberuli. Pedunculus sub flore modo 
0"25 cm. long., sat validus, puberulus, sub fructu incrassatus et usque ad 1'3 cm. 
elongatus; pedicelli pedunculo dimidio longiores, puberuli. Calycis lobi circa 
0"25 cm. long., nervosi. Ploris nondum expansi petala 0'28 cm. long., ungue vix 
O'l cm. long. Tubus stamineus ejusdem floris circa usque ad \ divisus, et antherae 
circa 003 cm. long. Pistillum 0*1 cm. long., ovarium solum duplo excedens. 
Capsula 1"2 cm. long, et lat., minute tomentella, tubercula summa erecta, 0'3-0"5 
cm. long., inferiora patula vel decurva et pauUo breviora. Semina 0-7 cm. long., 
0*4 cm. lat., subtrigona, ferruginea, levia ; bilum albidum ; cbalaza ferruginea. 
Species nuUi mihi cognitse affinis, et ob fructum singularem facili obtutu cognoscenda. 

TILIACE^. 

CoiicHOB.rs ARGiTUs, H. B. K. Xov. Gen. et Sp. v. p. 262. 

Rah. Inveni juxta Corumba mens. Jan. (N. 1006.) [Bras. Or., Guiana, Nov. 
Granat.] 

Ltjhea tjnieloea, a. St.-Hil. PL Bras. Merid. i. p. 290. 

Rah. Ploret ad Santa Cruz mens. Nov. (N. 591.) [Bras. Or., Bolivia.] 



OF THE MATTO GEOSSO EXPEDITION, 1891-1)2. 321 

LrHEA SPECIOSA, Willd. in Neue Sclir. Gesell. nat. Fr. Berlin, iii. p. 409. 

Hah. Crescit in cacumine montium Serra da Chapada, ubi tloret mens. Aug. (N. 139.) 
[Amer. Trop., Cuba.] 

Sloanea Maximowicziana (?), K. Schum. in Mart. Fl. Bras. xii. pars iii. p. 192. 
Hah. Viget in ripa fl. Paraguay inter Santa Cruz et Diamantino. (N. 614.) 
The specimens are without flowers, hence I am unable to determine the siwcies with 

certainty. If not ^S*. Maxlmoioicziana, we have here a new species closely alli(^d to it. 

[Bras. Or.] 

LINACE.E. 

Erythroxylox anguifugum, Mart, in Abhandl. k. bayer. Akad. Wiss. iii. p. 3(52. 

Hab. Crescit in ripA. fl. Paraguay, baud procul a Tres Barras, mens. Oct. florens. 
(N. 620.) [Matto Grosso.] 

Erythroxylox dapiinites, Mart, in Abhandl. k. haver. Akad. Wiss. iii. p. 363. 
Hab. Beperi in angustiis ad Serra da Chapada, ubi floret mens. Aug. (N. 77.) 
After spending much time over this plant, I come to the conclusion that the above is 

its correct name. Mine is a form with long stamens, much longer than in the figures of 

this species in Mart. El. Bras. xii. pars iii. tab. 26. [Goyaz.] 

Erythroxylon nitidum, Spreng. Syst. Veg. ii. p. 390. 

Hab. lleperi ad Santa Cruz mens. Nov. (N. 452.) [Bras. Or., Guiana.] 

Erythroxylon campestre, a. St.-Hil. El. Bras. Mend. ii. p. 97. 

Hab. Legi aut ad Cuyaba aut ad Serra da Chapada mens. Aug. (N. 151.) 

By an oversight I omitted to note at the time the habitat of this plant. [Bras. Or., 

Goyaz, Guiana.] 

Erythroxylon precox (sp. nov.) ; ramulis gracilibus, brachycladis, ramentis arete 
imbricatis instructis ; foliis juvenilibus oblongo-ovatis, basi angustatis, apice mucro- 
natis, adultis verisimiliter membranaceis, subtus plicis destitutis; stipulis petioles 
paullo excedentibus, crebre striatis, dorso primo ferrugineo-tomentosis, dein glabris, 
floribus praecocibus, pedicellatis ; sepalis parvis 0-12 cm. long. ; petalis vix 0'3 cm. 
long. ; drupa .... 
Hab. Crescit in sylva primgeva ad Santa Cruz, ubi floret mens. Sept. (N. 341.) 

Erutex paullo ultrametralis, deorsum aphyllus. Rami 0-3-0-4 cm. diam., teretes 
cortice cinereo longitrorsum rimoso obtecti, ramuli brachycladigeri crecto-ascen- 
dentes, circa O'l cm. diam. Kovelli compressi. Eolia vix omnino evoluta 2-0 cm. 
long., 1'2 cm. lat., glabra, subtus pallida, adulta desiderantur ; costa media utrinque 
prominula, costse secundariae plures, debiles, angulo fere recto insertae ; venulae arete 
reticulata, supra vix aspectabiles ; petioli (an omnino evoluti ?) 0*2 cm. long. 
Stipule oblonga", dorso carinatae, apice breviter bisetuliferse, striis pluribus in- 
structse, subscariosse, 0'25 cm. long. Bamenta triangulari-lanceolata, inferiora 
obtusa, dorso brevissime aristata vel nuda, superiora acuminata et basi arista 
dorsali circa O'l cm, long, munita, omnia carinata, crebre striata, membranacea. 
ProphyUa 0*1 cm. long., spathulato-oblonga, superne cilia to-erosula ibique dorso 
pilifera. Pedicelli 0-3-0-5 cm. long., sursum dilatati. Sepala basi connata, tri- 

SECOXD series. — BOTANY, VOL. IV. 2 U 



322 MR. SPENCER LE M. MOORE — PHANEROGAMIC BOTANY 

angulari-oblonga, acuta. Petala alba, oblonga, obtusa, ligula duplicata; lobi laterales 
antici, late ovati. lobo intermedio minimo lancelato rigide acuminato multo lon- 
giores ; lobi laterales postici anticos semiseqnantes, connati, apice inflexi. Tubus 
stamineus calyci subsequilongus, ore breviter lobulatus ; filamenta subsequilonga, 
tubum stamineum et stylos florum brachystylorum tantum scrutatorum circa 2-plo 
excedentia. Ovarium oblongo-ovoideum, tubum stamineum gequans ; styli 3, 
discreti ; stigmata clavellato-capitata. 
Species fortasse juxta Erythroxylon m/yrsimte. Mart,, interponenda, ab eo attamen 
multis signis gravioris momenti insigniter abborret. \Bras. Or.] 

Erytheoxylon dtjeum (sp. nov.). Fruticosa, ramosa, glabra, ramulis teretibus, sat crassis, 
in longitudinem eximie rimatis, cinereis, foliis brevipetiolatis, majusculis vel sub- 
mediocribus, oblongo-obovatis,obtusis vel obtusissimis, basin versus sensim angustatis, 
rigide coriaceis ; stipulis elongatis, lineari-lanceolatis, basi amplexicaulibus, petiolo 
longioribus ; ramentis stipulis conformibus, apice longiuscule apiculatis, superioribus 
dorso prope basin aristatis ; sepalis sub antbesi 0'25 cm. long,, sub fructu paullulum 
(usque 0*3 cm.) auctis, ovatis, breviter acuminatis, medio nervis parallelis approximatis 
percursis, petalis vix 0*4 cm. long, 
JELah. Heperi in sylva primseva prope Santa Cruz, necnon in ripa fl. Brasinbo ; floret 
menss, Sept. et Oct. (Nn. 312, 500.) 

Frutex diffusus, ramosus. Eami obsoleto-puberuli, mox glabri. Foliorum lamina 
usque 17*5 X 9-0 cm., modiea 11-0x6*5 cm., supra nitida, subtus pallida; costa 
media supra impressa, subtus admodum eminens ; costse secundarise utrinque circa 
12, ssepius plus minus undulatse, dein subito fornicatim-conjunctse, subtus prominulse ; 
venulse laxe reticulatce, parum eminentes ; petioli circa 0*5 cm. long., sat obesi, 
supra late canaliculati. Stipula3 sagittato-lanceolatse, dorso bicarinatse, longitrorsum 
striatse, membranacese, 012 cm. long., apice setis tribus 0"12-0-2 cm. long, instructse, 
castanese. Ramenta mox laxiuscula, inferiora superioribus multo breviora, ilia 
triangularia, hgec lanceolata et apice tri-setulifera, inferiora dorso breviter aristata 
vel nuda, superiora basi arista vix usque 0*2 cm. long, onusta. Propbylla conferta, 
ovato-lanceolata, acuminata, subscariosa, Plores nodis bracteatis plurifloris insidentes, 
et pediccUis ±1*0 cm. long, quadrangularibus fulti. Sepala basi connata, in antbesi 
0-2 cm. lat. Petala alba, oblonga, obtusa, circa 0-18 cm. lat, ; ligula duplicata, 
petali lamina circa duplo brevier; lobi laterales antici concavi, ovati, obsolete 
undulati, lobum intermedium oblongum bifidum bene superantes; lobi laterales 
postici anticos semia^quantes, connati, apice inflexi. Tubus stamineus circa 0-12 cm, 
long., i. e. calycem semia^quans, ore undulato. Plorum dolicbostylorum a me 
tantummodo scrutatorum filamenta ejusdem seriei insoquilonga, breviora tubum 
stamineum ^-plo excedentia, et filamenta longiora saepissime f a3quantia. Ovarium 
subglobosum, tubum stamineum pauUo excedens ; styli 3, discreti ; stigmata capitata. 
Drupa ovoideo-oblonga, 07 cm. long. 
Ex affinitate E. macrophijlU, Mart., abs qua facili obtutu secernendum ob folia minora 
diversiformia, stipulas longius setuliferas, inflorescentiam disparem, sepala breviora 
ovata nequaquam flabellatim-nervosa, tubum stamineum calycem semia^quantem neque 
eo multo breviorem, drupam fere duplo minorem, etc. FAmazoAia, Guiana.] 



OF THE MATTO GEOSSO EXPEDITIOX, 1801-92. 32:? 

MALPIGHIACE.E. 
Byrsonima VEEBASCiroLiA, Ricli. in Ann. Mus. Paris, xviii. p. 181. 

Bab. Ad Serra da Clmpada satis abundat, mens. Aui?. florcns. (Nn. 151. 155.) fAmer 
Austr. Trop.] » -^ L • 

Byksonima cydoni^folia, A. Juss. in A. St.-Hil. Fl. Bras. Merid. iii. p. 77. 

Bah. Crescit ad Santa Cruz, ubi floret menss. Sept.-Nov. et Dec. fructificat. 

This seems to be the rar. ch'quUrmh of Jussieu's Monograph (p. 27). 

A very common small tree, of xvhieh, by some oversight, I failed to obtain flowering 
specimens. 

Byrsonima cydoni^folia, A. Juss., var. cuneata, S. Moore, a typo abhorrens foliis 
lanceolatis deorsum cuneatis, peduneuUs et pedicellis gracilibus, his elongatis. An 
potius species propria ? 
Bab. Pruticulus f -metralis ; floret ad Jangada mens. Sept. (N. 274.) [Bms. Or., 

BoUvia.] 

Byrsonima coccolob.^folia, H. B. K. Nov. Gen. v. p. 148. 

Bab. Ad Santa Cruz flores albos ostendit mens. Nov. (N. 782.) [Amer. Austr. 
Trop.] 

Byrsonima indorum (sp. nov.). Arbuscula, abunde ramosa; foliis brevipetiolatis, 
anguste obovato-oblongis, raro obovatis, obtusis, nunc brevissimc et obtuse cuspidatis, 
basi acuminatis, tenuiter coriaccis, appresse pilosulis, fere glabris ; racemis termi- 
nalibus paucifloris, subferrugineo-tomentosis ; bracteis bracteolisque approximatis, 
illis triangulari-oblongis ; pedicellis J-plo brevioribiis has rotundato-ovatas 2-plo exce- 
dentibus; petalis flavis; antherarum sagittatarum loculis ultra connectivum obtusum 
brevissime productis una cum ovario piliferis. 
Bab. Beperi in ripa fi. dos Bugres mens. Oct. florentem. (N. 501.) 

Kami usque 0*5 cm. diam., teretes, cinerei, rimosi, et Icnticelliferi ; ramuli appresse 
hirsutuli, cito glabri. Poliorum lamina supra hand nitida, subtus paullulum decolor, 
modica 100 cm. long., 4'5-6"0 cm. lat. ; costae secundaria) utrinque 8-10 angulis 
fere rectis insertse, leviter arcuata?, marginem versus undulatim et arcuatim con- 
junctse, una cum costa media supra parum subtus maxime eminentcs ; venula) arete 
reticulatge utrinque prominulse ; petioli usque I'O cm. long., appresse hirsutuli. 
Racemi 3'0-7'0 cm. long. Bractea? apice acutic, dorso appresse puberula^, 02 cm. 
long. Bracteolae O'l cm. long., obtus^e, puberulai. Pedicelli OS cm. long., patuli, 
subferruo-ineo-tomentosi. Flores expausi (ex schedis meis flavi) desiderantur. 
Calycis 10-glandulosi lobi ovati, obtusi, puberuli, margine ciliolati, 0'28 cm. long. ; 
glandulie 012 cm. long., oblongo-obovat^e. Petola extus glabra. Pilamenta bre- 
vissima, dense pilifera ; antherae 0-2 cm. long., O'OG cm. lat. Ovarium triangularis 
ovoideum, vix O'l cm. long. ; styli crassiusculi, glabii, paullo ultra 0-2 cm. long., 
apice incurvi. Fructus ignotus. 
Forsan cum Byrsonima intermedia, A. Juss.. et B. Icevigata, DC, componenda, ab 

2u 2 



324 MB. SPENCER LE M. MOOEE — PHAXEEOGAMTC BOTANY 

ambabus revera sat discrepat ob folia necnon bracteas dissimiles. Illius antberse insuper 
apice cuspidato-acuminatai, Imjus antherse Decnon ovarium glabrum, etc. [Bras. Or,, 
Amazonia.] 

Thryallis Labtjrntjm (sp. nov.). Prutex volubilis, caule sat tenui, subtereti, tenuiter 

sublutescenti-tomentoso, abnnde ramuloso ; foliis amplis, petiolatis, late ovatis, obtusis 

vel acutis, basi rotundatis, supra mox (prseter nervos) glabris, subtus tenuiter albo- 

tomentosis, chartaceo-membranaceis, in sicco virescentibus ; fioribus ultra 2-0 cm. 

diam. ex axillis foliorum superiorum ortis et in racemis pauci- vel plurifloris foliis 

liebetatis sen validis, intermixtis, digestis, aut paniculas terminales plurifloras 

patulas conficientibus ; calycis ampli tenuiter tomentosi segmentis insequalibus, ovatis 

vel ovato-oblongis, obtusis ; petalis insequalibus, longe unguiculatis, rotundatis vel 

rotund ato-reniformibus, margine erosulis; staminibus calyce duplo brevioribus; 

fructibus magnis triquetris, fere 0*5 cm, diam. 

Hcib. Crescit in nemorosis prope Corumba, mens. Jan. florescens. (N. 970.) Incolit 

itaque Brasiliam Orientalem, scilicet provinciam S. Catharinam, unde raisit Tweedie (ex 

spec, in Hb. Kew.) et Eritz Miiller (n. 188 in Hb, Kew.). 

Ramorum floriferorum caulis 2-0-3"0 cm. diam., demum glaber et subcinereus et in 
longitudinem ramosus. Eoliorum pagina nonnunquam basi parum obliqua, ad ll'O cm- 
long, (plerumque vero 8'0-10'0 cm.) et 5"0-6'0 cm, lat. (raro 8'0 cm.), margine obscure 
undulata, subtus ima basi (sub apice petioli ipsius) glandulas duas elevatas tomentosas 
rotundas vel parum ellipticas 0*1 cm. diam. prsestans ; costae secundariae utrinque 
plerumque 5-8 angulis latis insertse, primo rectse, marginem versus subito arcuatim- 
conjunctifi, una cum costa media supra impressa subtus eminentes; venulae laxe 
reticulatse, inconspicuae ; petioli 0*8-l"7 cm. long., pingues, breviter canaliculati, ima basi 
subito incrassati, molliter tomentosi. Inflorescentiae axis molliter tomentosa; racemi 
foliis breviores, paniculae pleraeque folia superantes ; bractese caducissimce, baud visae ; 
pedicelli graciles, tomentosi, 0*7-l"0 cm. long. Flores 2-5 cm. diam. Calycis segmenta 
ad 0-85 cm. long., utrinque tenuiter tomentosa, obscure nervosa. Petala lutea, vix I'O 
cm. long., usque ad 08 cm. lat., eorum unguis ad 0'6 cm. long., tenuis vel firmior ac 
(saltem in sicco) fuscior, Antherae oblongae, filamentis compressis aequilongae. Ovarium 
trigonum, tomentosum, 0*2 cm. diam. ; styli glabri, stamina sequantes. Calycis persis- 
tentis segmenta sub fructu firma, leviter tomentosa, circa 1*0 cm. long., 0-5-0'6 cm. lat., 
plana, patentia, intus manifesto reticulato-nervosa ; fructus tomentosi crista dorsali 
valde prominente. 

Species distinctissima et facile recognoscenda. 

Recourse lias been had to Tweedie's specimen for the description of the fruit of this 
"beautiful plant. 

Hujus stirpis reperi ad Corumba mens. Jan. (n. 969) varietatem insequentem : — 

Var. MiNou, a typo distans habitu (frutex 3 met. alt.) baud volubili (ascendente, 
an nonnunquam scandente ?), ramulis gracilioribus, foliis minoribus (ad 6-5 cm. long., 
plerumque vero brevioribus), 2-0-3-5 cm. lat., fioribus minoribus (2'0 cm. diam.), 
calycis lobis et petalis minoribus. 



OF THE MATTO GROSSO EXPEDITIOX, 1891-92. 325 

A small specimen in the Herbarium at Kew, collected by Blauchet (no. 3028), is 
perhaps to be referred to this variety, thougii its thicker leaves are stellately pubescent on 
their underside, not tomentose. [Bras. Or.] 

Galphimia brasiliensis, a. Juss. in A. St.-Hil. Fl. Bras. Merid. iii.p. 71. 

JSab. Floret ad Jangada mens. Sept. (N. 288.) [Bras. Or., Goyaz, Amazonia.] 

Heteropteris (§ Ptycheteropteris ?) NTjDiCAULis (sp. nov.) ; caule erecto, raraoso, 
tereti, ruf o-tomentoso, demum glabro ; f oliis breviter petiolaiis, oblongis, obtusiusculis 
vel acutis, plerisque brevissime acuminatis, membrauaceo-coriaceis, su]]ra appresse 
pubescentibus, demum glabris et pallide nitentibus, subtus molliter rufo-tomentosis, 
prope basin brevissime cordatam biglandulosis ; umbellis paucifloris (plerumque 2-i- 
floris) paniculas breves axillares vel terminales foliis subnequilongas efformantibus, 
et saepe foliis hebetatis mixtis rufo-tomentosis; prophyllis maxime ina^qualibus, ovatis, 
basi biglandulosis, rufo-tomentosis; bracteis parvis bracteolas a^quantibus, pediccllis 
tenuibus, nequaquam incrassatis, juxta basin articulatis ibidemque bibracteolatis ; 
calycis 8-glandulosi segmentis late oblongis, obtusis, glandulas duple excedentibus, 
superne paullo incurvis; petalis ovatis obtusissimis, longiuscule ac crassiuscule ungui- 
culatis ; samaris rufo-tomentosis lateribus gibbis planis, alis acquimagiiis vel sub- 
aequimagnis vel inaequalibus, necnon interdum alfi una solummodo evoluta. 
Hub. In fruticetis reperi ad Cuyaba (n. 16) et ad Santa Cruz (Barra dos Bugres) 
(n. 335). Floret mens. Aug. ; fructus (vix matures ?) prssbet mens. Sept. 
Frutex gracilis, 1-vix 2 met. alt. e rhizomate pingui erectus, basi aphyllus, saepe post foli- 
orum casum florens. Folia 3" 5-8-0 cm. long, (raro ad apices ramulorum folia minora 
posita sunt), 1-5-3-5 cm. lat., saepe subtus ad sinistrum latus glandula tertia aucta; 
costse secundarige utrinque 5-6, leviter arcuatse, sub margine conjunctoe, subtus 
eminentes; venulas laxe reticulatae; petioli 0*2- 0-4 cm. long., crassiusculi, rufo- 
tomentosi. Paniculse plerajque 30-4-0 cm. long, (occurrunt vero ad 80 cm. long.), 
subcongestse ; prophylla ad 01 cm. long, (ac igitur revera in folia hebetata trans- 
euntia), pleraque vero circa 0-2 cm. long. Pedicelli 0'7-l-2 cm. long., subcongesti ; 
bracteae bracteolseque oblongo-lanceolatse, rufo-tomentoste, circa 0'15 cm. long. 
Flores lutei, I'O cm. diam. Calycis segmenta 0*35 cm. long. ; glandulae 016 cm. 
long., anguste ellipticae ; segmentum quintum lineas duas elevatas pro glandulis 
ostendens. Petala 0-6 cm. long., margine undulata ; unguis vix 02 cm. long. 
Ovarium leviter triquetrum, rufo-tomentosum. Samara (vix matura) I'S cm. lat., 
0-6 cm. long. 
Species forsitan juxta S. confertifloram, A. Juss., i)onenda, sed multis notis dispar. 
The Cuyaba specimens are in flower, those from Santa Cruz in the fruiting stage, and 
except for the former having somewhat larger leaves, less rusty beneath, I can see no 
difference between them. [Bras. Or.] 

Verisimiliter in hoc genus digerenda et juxta speciem prsecedentem interponenda exstat 
planta fruticosa metralis quam ad Santa Cruz legi (n. 336), caule erecto, tereti, superne 



326 ME. SPENCER LE M. MOOEE— PHANEROGAMIC BOTANY 

taiitimi foliigero ; foliis minimis, subsessilibus, numniul arils, apiciilatis, margine revolutis, 
coriaceis, glabris, 0-8-1-0 cm. long, et lat. ; umbellis paucifloris (4), fermgineo-pube- 
scentibus, in paniculas effusas folia multo excedentes digestis; bracteis parvis, lanceolatis, 
feiTugineis, circa 01 cm. long. ; pedicellis gracilibus, centimetralibus vel ultra, ferrugineo- 
pubescentibus, pauUo infra medium articulatis et bracteolis caducis onustis; calycis 8- 
glandnlosi segmentis OYato-oblongis, dorso ferrugineis, 0*23 cm. long. ; glandulis oblongis, 
()"12 cm. long., filamentis glabris, ovario tomentoso, fructu ignoto. 

Ad Serra da Chapada mens. Ang. (n. 81) legi specimen forsan in boc genus tra 
hcndum, foliis lanceolatis acuminatis, basi obtusis, glabris, coriaceis, supra nitidis, subtus 
pallidis, ima basi glandulis duabus circularibus 0-04 cm. diam. instructis, IOO-17'O cm. 
long, et 3'5-5-5 cm. lat. ; costis secundariis utrinque circa 10, angulis latis insertis, aperte 
arcu.atis, una cum costa media subtus conspicuis ; venulis arete reticulatis, supra admodum 
aspectabilibus ; petiolis incrassatis, transverse rugatis, O'o-O'S cm. long. ; j^aniculis termi- 
nalibus, quam folia brevioribus, laxe plurifloris, appresse strigoso-joubescentibus ; bracteis 
bracteolisque parvis, subulatis, l-0-l"5 cm, long. ; pedicellis elongatis, gracilibus, juxta 
basin articulatis, puberulis, O'l cm. long.; floribusluteis; calycis 8-glandulosi lobis oblongis, 
obtusis, 0*2 cm. long., glandulis oblongis 0*15 cm. long. ; petalorum lamina 0*5 cm. long., 
0'3 cm. lat., et ungue O'l cm. long. ; antberarum dorso strigoso-pubescentium connectivo 
apice mucronato, loculis 0'25 cm. long., filamentis strigose i)ubescentibus 0*12 cm. long., 
ovario sericeo, samara ignota. 

Banisteria pubipetala, a. Juss. in A. St.-Hil. Tl. Eras. Merid. iii. p. 41. 

Ad Santa Cruz flores luteos ostendat mens. Dec. (n. 792), itaque ex Matto Grosso a 
cl. Leeson missa (n. 11 in Herb. Mus. Erit.). [Bras. Or., Goyaz, Amazonia.] 

Banisteria constricta, Griseb. in Mart. El. Bras. xii. pars i. p. 46. 

Hah. Juxta Cuyaba et ed Serra da Chapada satis abundat, mens. Aug. florens et 
fructificans. (iSTn. 15, 104.) [Amazonia.] 

Peixotoa hirta, Mart, ex A. Juss. Monog. Malp. p. 176. 

Hab. Legi ad Serra da Chapada mens. Aug. (N. 153.) 

Samarae sericeo-tomentellae, virides, alse late oblongse, apice truncate et undulatse,. 
marginibus ambobus rectis vel fere rectis, 2*0 cm. long., 1-2-1-5 cm. lat., crebre nervosa?. 
[Bras. Or., Goyaz.] 

Stigmaphyllon calcaratum, N. E. Br. [in Trans. Bot. Soc. Edinb. xx. (1894) 48]. 

Hah, Viget in paludibus juxta Corumba, mens. Jan. florens. (N. 1012.) [Para- 
guaria.] 

Tetrapteris (§Pextapteris) pilifera (sp. nov.); caule ramoso, ramulis subteretibus, 
ad nodos parum tumidis, viUoso-tomentosis, demum glabris ; foliis subsessilibus, 
oblongo-ovatis, obtusis vel acutis vel brevissime cuspidato-acuminatis, supra pilosis, 
subtus lanato-villosis, vix tomentosis ; corymbis 4-floris, quam folia brevioribus ; 
pedicellis gracilibus, medio articulatis ibidemque bibracteolatis ; bracteis lineari- 



OF THE MATTO GROSSO EXPEDTTIOX, 1891-92. 327 

lanceolatis, bracteolas excedentibus ; calycis villosi se^mcntis Lmcoolato-obloni^is, 
obtusiusculis, sub friictu glandulas phis qiiam 2-plo excedentibus ; samara3 alis late- 
ralibus 4, discrctis, inaequalibus, nunc late obovatis, nunc oblonij^o-spathulatis, margine 
eroso-undulatis ; ala dorsali subquadrata, crosa, quam laterales breviore. 
Sab. Vigetin nemorosis ad Santa Cruz, mens. Nov. fructum proebens. (N. 502.) 

Fmtex incertse altitudinis, copiose ramosus. Polia ad 80 cm. long., medio ad 33 cm. 
lat. basin versus leviter coarctata, margine undulatii et basin A'crsus passim glandu- 
lifera, raembranacea subtils aliquanto decoloria ; nervus medius pag. suporiore 
appresse tomentosus pag. inferiore eniinens ; costaB secundarirc utrinque circii 0, 
oblique insertse, erecto-ascendentes, paruni aspectabiles ; venulce subobsoletae. 
Petiolus 0" 2-0-25 cm. long, lanato-villosus, ad apicem faciei superioris glandulis 
duabus rubellis parvis (vix 0"05 cm. long.) gaudens. Pedicelli 1'1-1'3 cm. long., 
lanato-villosi. Bractese liucari-laneeolatse, dorso lanato-villosa?, inferne glandiiloso- 
bidenticulatae, 03 cm, long. Bracteolce bracteis similes nisi minorcs, 0*2 cm. long. 
Plores desunt. Calycis segmenta sub fructu 0"i cm. long.; glandula) 0'2 cm. 
long., ellipticse, superne coarctatse. Pructus circa 2*5 cm. diani. Samara? rubclbe aloe 
laterales majores 0*9 cm. long., 08 cm. lat. ; minores majoribus icquilongi, 0'5 cm. 
lat. ; ala dorsalis 0'5 cm. long., 06 cm. lat. ; alai omnes papyracege, eximie 
nervatse. 
Ex affinitate Tetrapteridis metalUcoloris, A. Juss., sed aliquot notis dispar. 
I have ventured to describe this plant, although the only specimen which has survived 

is without flowers. A plant oC Burchell's collecting (n. 5754 in Herb. Kew.) should 

perhaps be referred to this species. [Bras. Or.] 

Teteapteeis (§ Pentapteris) precox (sp. nov.) ; caule ramoso, tereti, lanato-tomentoso, 
demum glabro, cortice fusco-cinereo crebre lenticellifero obtecto ; foliis petiolatis (ut 
apparet) ovatis seu ovato-oblongis, margine obscure undulatis, chartaceis, supra 
glabris, subtus tomentoso-pilosis ; ramis StEpe umbellis prsecocibus instructis, umbellis 
lateralibus j)lerumque 2-6-floris ; pedunculis srepissime subsessilibus, prophyllis 
pluribus parvis linearibus sericeis instructis; pedicellis gracilibus, pedunculo 
multoties longioribus, supra medium articulatis et juxta medium bibracteolatis ; 
bracteis parvis linearibus, bracteolis conformibus ; calycis sericeo-velutini segmentis 
lanceolato-ovatis, quam glandulae vix duplo longioribus ; petalis ina^qualibus, ungui- 
culatis, ovatis vel panduriformibus, apice retusis ; samarae alis lateralibus omnino 
disjunctis, subsequalibus, late linearibus, retusis vel margine apicali eroso-undulatis, 
ala dorsali lateralibus subsimili, sed plus quam duplo breviore. 
Scih. Juxta urb. Cuyaba abundat, et mens. Aug. floros necnon fructus vix maturos 

praebet. (N. 17.) 

Prutex prolixus, circa 1 met. alt., inferne apbyllus. Polia, ut videtur, ad circa 5-5 cm. 
long, et 8-0 cm. lat. ; petiolus ad 0-6 cm. long., lanato-tomentosus, sat pinguis, 
glandulas 3 parvas in sicco atras et nitentes circa 0-015 cm. long, gerens. Pedunculi 
raro ad O'S cm. long., recti, sericei ; prophylla vix ultra 0-15-0-2 cm. long. Bractese 
usque ad 0-38 cm. long., una cum bracteolis 0'2 cm. long, sericese. Pedicelli l-O-^ix 



328 ME. SPENCER LE M. MOOEE— PHAIS^EROGAMIC BOTANY 

1-5 cm. long., subcortgesti, sericei; eoriim bracte{3e 0-2-0-3 cm. long. Elores Intei, 
circiter 1-0 cm. diam. Calycis segmenta acutmscula, 0*4 cm. long. ; glanduhe 
anguste ellipticse, marghie rugatse, 0-2-0-25 cm. long. Petala 0-5-0-7 cm. long., 
eoriim imguis ad 015 cm. long., nunc crassus, nunc tenuis. Ovarium sericeum, sub- 
globosum, circa 0-22 cm. diam. ; styli crassi, puberuli. Samarse sericeo-tomentosse, 
vix matura;, alge laterales ad 0"7 cm. long., 025 cm. lat. ; ala dorsalis 0-3 cm. long. ; 
omnes fere omnino glabra?, virescentes, papyraceae, in longitudinem crebre nervosa^. 
Cum Tetrapteridi ramiflord, A. Juss., comparanda, specie mihi descriptione tantum 

cognita, sed ob pedunculos copiose prophyllis instructos, pedicellos elongates, etc., certe 

distincta. 

This is one of the commonest plants round Cuyaba, but I bad to search some time 

before finding flowering shoots still bearing leaves, and then the latter were in a very 

imperfect state. [Bras. Or.] 

HiE^EA (§Mascagnia, Pleuropteris) kitexs (sp. nov.); caule terete, cinereo vel 
fiisco-cinereo, glabro ; foliis breviter petiolatis, ovatis vel ovato-oblongis, obtasis vel 
obtuse acutis, nonnunquam leviter cuspidatis, basi rotundatis vel levissime coarctatis, 
coriaceis, glabris, nitentibus, eglandulosis ; floribus breviter racemosis vel paniculatis, 
paniculis congestis, subthyroideis ; prophyllis minutis ; pedicellis infra medium 
minute bibracteolatis ; bracteis bracteolis similibus ; calycis 8-glandulosi segmentis 
late oblongis, obtusiusculis, quam glandular baud duplo longioribus ; samaris (an 
maturis?) parvis, dorso eximie carinatis et argute scrobicidatis, alis oblongis, 
mar2:ine inciso-undulatis. 
Hob. In convalle ad Serra da Chapada mens, Aug. fructus gerens. (N. 74.) 

Prutex scandens, a basi foliatus. Polia ad 15-0 cm. long, (plerumque S-O-ll'O cm.), 
4-0-6-0 cm. lat., reticulato-nervosa ; petioli 0-5-0-8 cm. long., crassi, juxta medium 
glandulas duas parvas ellipticas in sicco fuscas vix 0*2 cm. long, prcebentes. Costae 
secundarise circa 8, angulis latis plerumque insertae (inferiores raro erecto-ascen- 
dentes), aperte arcuataj, deinde margine parallelae et maxime undulatse ; venulae 
arete reticulatae, utrinque conspicuae. Inflorescentiae axis appresse pubescens, 
demum glaber. Panicula terminalis ad 5-0 cm. long., interne vix 5-0 cm. diarn. 
Prophylla et bractea? subulatae, ilia 0'15 cm., hae 0-1 cm. long. Pedicelli O'8-l-O 
cm. long., pubescentes. Calycis segmenta rufulo-tomentosa, vix 0*3 cm. long., 
glanduloe vix 0-2 cm. long., anguste ellipticae. Petala baud visa. Ovarium sub- 
trigonum, tomentosum ; styli glabri, ovario breviores. Samarae (an maturae ?) 0'7 
cm. long, et lat., membranaceae, virescentes. 
Ex afi&nitate K. cJilorocarpcdy A. Juss., abs qua ob foliorum formam et magnitudinem 

pluria inter propria eximie discrepat ; videtur itaque ad H. hippocrateoidem. Planch, et 
Triana, accedere, sed foliis latioribus, petiolis biglandulosis, inflorescentia saipe racerai- 
formi, bracteis minutis, etc., certe ab ilia di versa. [Bras. Or., Nov. Granat.] 

Hiii^A (§ Mascagnia, EriiASCAGNiA) voLUBiLis (sp. uov.). Frutcx volubilis, caulibus 
teretibus, pubesceutibus, demum fere glal^ris, in longitudinem rimosis; ramulis abunde 
foliosis, rufo-tomentosis ; foliis petiolatis, obtusis vel acutis vel breviter acuminatis. 



OF THE MATTO GKOSSO EXPEDITION, 1S91-92. 329- 

basi rotundatis, nervo medio excurrente apiculatis, supra piloso-pubesceutibus, 
demum prgeter ad nervos glabris, subtuspiloso-pubescciitibus, meinbranaccis, eglantlu- 
losis ; petioHs rufo-tomentosis, utrinque basi glandula auctis ; inllorescentiaaxillari vol 
terminali, corymbosa, vcl floribus in paniculas paucifloras corymbiformes dispositis ; 
pedicellis gracilibus, juxta medium minute bibracteatis ; calycis 8-glandulosi 
segmentis glaudulas baud superantibus; petalis ovato-oblongis, breviter unguieulatis; 
samaris albo-tomentosis, immaturis, ala dorsali parvi'i, latcrali samaram cingente 
amplfi suborbiculari. 
Hab. In nemorosis yiget ad Santa Cruz, menss. Sept.-Nov. florens. (Nn. 340, G47.) 

Caulis 2"5-3-0 cm. crass., purpureo-bruneus. Folia pleraque 7*5-9"0 cm. long, (raro ad 
lO'O cm.), plerumque 4*0-7"0 cm. lat., margine ciliata, pagiua super, ad nervum 
medium plus minus appressc pubesccnti-tomentosa, pagina infer, pallida, ncrvis 
aliquanto elevatis plus minus pubescenti-tomentosis ; petioli 0"7-l"5 cm. long., 
pingues, glandulis late ellipticis vel fere orbicularibus O'l cm. diam. Pedunculi rufo- 
tomentosi, plerumque l*0-30 cm. long. Propliylla subulata, rufo-tomentosji, circa 
0*2 cm. long. Pedicelli plerique circa I'O cm. long., subcongesti, rufo-tomentosi, 
bractea^ prophyllis subsimiles. Elores rosei, circa 1*2 cm. diam. Calycis segriienta 
0'3 cm. long., ovata, obtusa, pubescentia; glandular elliptical, calycis segmentis 
sequilongse. Petala 06 cm. long, et 0-45 cm. lat., nervo medio in sicco eminente ; 
uQguis modo O'l cm. long., sat pinguis. Ovarium trigonum, albo-tomentosum ; 
styli vix omnino glabri. Samarse immaturae circa 0*6 cm. long, et 05 cm. lat., ala 
dorsali circa 0*1 cm. alt. 
Species accedens ad Hir(Eam ruhram {Maseagniam, Griseb.), qua) forsan ad scctionem 

Bumascagmam transferri debet, sed foliorum crassitudine ac vestitu, corymbis brevibus 

nee solis terminantibus, pedicellis brevioribus, bracteis omnino diversis, etc., optime 

divergens. 

HinJEA SEPIUM, A. Juss. in A. St.-Hil. PI. Bras. Merid. iii. p. 19. 

Hab. Ad Santa Cruz flores et fructum siniul ostendit mens. Nov. (N. 520.) 

HiR^A SEPIUM, A. Juss., var. nitens, S. Moore, foliis quam ea typi firmioribus supra 
nitidis et puberulis, samaris ignotis. 

Bab. Ad Corumba floret mens. Jan. (N. 1008.) 

Qnum samaras non examinavissem, exstitit qusedam quaestio num mera pro varietate 
haberi debeat. [Bras. Or.] 

HiRyEA cuJABENSis ?, Griseb. in Mart. PL Bras. xii. pars i. p. 100. 

Hab. Crescit in loco ignoto, verisimilter ad Santa Cruz. 

The specimen is very incomplete, and, owing to the loss of the label accompanying it, 
I am unable to give its locality. [Matto Grosso.] 

HiR^A (§Mascagnia) (sp. nov.) ; foliis et floribus ignotis ; samaris magnis, papyraceis, 
glabris, viridibus, ala dorsali 2*0 cm. long, et usque ad IS cm. lat., alis lateralibus late 
obovatis, truncatis, margine superiore dentato-undulatis, 30 cm. long, et lat. 

SECOND SERIES. BOTANY, VOL. IV. 2 X 



.•330 ME. SPEXCEE LE M. MOOEE — PHAXEEOGAMIC BOTANY 

Hah. ]VIatto Grosso [Leeson, in Herb. Brit. Mus.). 

A fine species, with fruits much like those of THrma glaucescens, Griseb., but much 
larger. Perhaps it is still nearer the Mexican H. macroptera, DC. 

Camarea ericoides, a. St.-HiL, PI. Kem. i. p. 158. 

Hah. Crescit ad Jangada, mens. Sept. florens. (N. 286.) [Bras. Or.] 

Specimina manca re vera fructibus carentia, quapropter quoad genus omnino dubia : — 

(i.) Volubilis, caulibus foliosis, strigose pubescentibus, cito glabris ; foliis oblongo-ovatis, 
obtusis obtusissimisve, nonnunquam retusis, basi rotundatis, supra glabris, baud 
nitentibus, subtus pilis strigosis arete appressis vestitis, tenuiter coriaceis, 5'0-10"0 
cm. long., 3'0-50 cm. lat.; petiolis 0-5-0'7 cm. long., appresse et strigose puberulis, 
basi biglandulosis insidentibus ; costis secundariis utrinque 8-10, primo fere rectis, 
dein dichotomis et arcuatim junctis ; paniculis terminalibus axillaribusve (et tunc ex 
axillis superioribus ortis), quam folia multo brevioribus (2"0-3*5 cm. long.), appresse 
ferrugineo-hirtulis, in umbellas brevipedunculatas plerumque S-l-floras digestis ; 
prophyllis lanceolatis, ferrugineo-hirtulis; bracteis bracteolisque approximatis, lanceo- 
latis, illis has duplo excedentibus et vix 02 cm. long. ; pedicellis 0"4-0*6 cm. long. ; 
calycis 8-glandulosi segmentis oblongis dorso ferrugineo-hirtis 0*25 cm. long., 
glandulis nunc minimis, nunc usque ad 0*16 cm. long., oblongis; petalis luteis, 
filamentis glabris, ovario tomentoso. 
Hah. Reperi juxta Corumba mens. Jan. florentem. (N. 945.) 

(ii.)P ruticosa ? ramosa, ramis subteretibus, strigose j^ubescentibus, mox glabris ; foliis 
ovatis, breviter acuminatis, rarius cuspidulatis, basi rotundatis, nunc utrinque glabris, 
supra baud nitidis, subtus pallidioribus, ima basi biglandulosis, 7'5-9"0x4-0-5'0 cm.; 
petiolis vix usque 1*0 cm. long., villosulis, cito glabris, inflorescentiis terminalibus axil- 
larisbusve ; iloribus nunc racemosis, nunc in umbellas paucas 2-4-floras brevipeduncu- 
latas paniculatim digestis, semper quam folia brevioribus nee ultra 6*0 cm. long. ; 
bracteis 0'25 cm. long., lanceolatis, pedicellos ferrugineo-hirtulos semisequantibus ; 
bracteolis sub flore insertis ovatis, obtusis, basi biglandulosis, 1-5 cm. long.; calycis 8- 
glandulosi segmentis oblongis, obtusis, dorso appresse hirtulis, 0'3 cm. long., glandulis 
oblongis subsequalibus usque ad 0-25 cm. long. ; filamentis glabris ; ovario tomentoso. 
Hab. Ad Corumba floret mens. Jan. (N. 1095.) 

(iii.) Eruticosa, deorsum aphylla, ramis erecto-ascendentibus, subteretibus, cinereo- 
purpurascentibus; foliis oblongis vel oblongo-lanceolatis, obtusis, basi parum coarctatis, 
utrinque (costa media subtus brevissime albido-sericeo-tomentella exempta) glabris, 
coriaceo-membranaceis, subtus nitentibus, costis secundariis utrinque circa 10 fere 
rectis, deinde margini parallelis, subtus conspicuis, costis tertii ordinis plurimis 
maxima pro parte secundariis plus minus perpendicularibus ; petiolis circa 0-5 
cm. long., sericeo-tomentellis, sursum glandulis duabus pedicellatis onustis ; floribus 
umbellatis, umbellis abbreviatis, 3-4-floris, I'o cm. long., sericeo-tomentellis ; l)racteis 
bracteolisque minutis, approximatis, sequalibus, ovatis, sericeis ; pedicellis 0*7 cm 



OF THE MATTO TrROSSO EXPEDITION, 181)1-1)2. 331 

lon^. et revera pedunculum vix 2-plo excedcntibus ; calycis 8-glaiululosi segmcntis 
ovatis, obtusis, dorso sericcis, 2-3 cm. long., glaiidulis ellipticis 01 em. long.; lilamentis 
pilosulis; ovario trilobo tomentoso. 
Hah. Verisimiliter invciii ad Santa Cruz, ob schedulam vero praitermissam locum 
accurate indicare nequi'o. 

GERANIACEyE. 
OxALis HiRSUTissiMA, Zucc. Monog. Anicr. Oxal. p. 179. 

Jlah. Ad Santa Cruz tloret mens. Nof. Norn, vernac. Azedhiha. (N. (508.) 
Prutex altus, interdum orgyalis. Plores lutei. Capsula3 viridcs, ovoidcje, angulis 
appresse hirtula3, 06 cm. long. Semina elliptica, 015 cm. long., transverso-rugata, 
brunea. [Bras. Or.] 

OxALis SEPIUM, A. St.-Hil. PI. Bras. Merid. i. p. Ill, var. picta. 

Mah. Crescit ad Jangada necnon ad Santa Cruz, monss. Scpt.-Nov. florcns. (Nn. 263, 
328, 561, 645.) [Bras. Or., Guiana, Ind. Oec] 

OxALis CATHARiNENSis, N. E. Br. in Gardn. Chron. 1887, I. p. 110. 

JELah. Crescit in sylviilis umbrosis juxta Corumba, ubi flon^s albos mens. Jan. priel)et. 
(N. 967.) 

My specimens diifer from the type (E. Mueller, no. 395, in H(u-l). Kew.) in liaviiig 
somewbat sliarper-angled leaves. In cultivation this beautiful little plant woefully 
deteriorates, possibly in consequence of exposure to light of too high intensity ; in its 
native vi'^oods, at least at Corumba, it prefers the darkest recesses. [Bras. Or.] 



RUTACE^. 

MoNNiERiA TRiFOLiA, Linn. Syst. Nat. ed. X. p. 1153. 

Hah. Crescit in sylva primseva inter Santa Cruz et Tapirapuan, itaque ad ripas fl. 
Paraguay inter Santa Cruz et Diamantino. (Nn. 396, 625.) [Bras. Or., Amazonia, 
Guiana, Nov. Granat.] 

SIMARUBACEJE ? 

Planta incertae sedis, forsan inter Simarubaceas dif&cile digerenda. Arbuscula, ramulis 
teretibus striatis, cortice cinerascente amaro cincta ; foliis alternis imparipinnatis, 
foliolis 8-10 alternatim insertis, oblongo-obovatis, basi longe et sensim cuneatis ; 
acuminatis, glabris, utrinque nitidis, creberrime et minute glanduloso-punctatis, 
spicis elongatis, exaxillaribus vel ex axillis foliis orbis ortis, simplicibus, raro 
bifurcatis ; calyce 5-partito ; petalis cum calycis lobis alternantibus, staminibus . . . ; 
ovario omnmo integro, 2-loculari, stigmatibus 2 brevibus sessilibus divaricatis coronato, 
ovulis quove in loculo 2, ab apice pendulis ; raphe ventrali, micropyle superiore ; 
baccis oblongis, 1-2-locularibus, loculis abortu l-spermis ; seminibus testa tenui arete 
adhserente obtectis, exalbuminosis ; cotyledonibus car no sis, radicula inconspicua. 

2x 2 



532 ME. SPENCER LE M. MOOEE— PHANEROGAMIC BOTANY 

Hab. Reperi in ripa fl. Paraguay inter Santa Cruz et Tres Barras, ubi floret mens. 

Oct. (N. 491.) 

Bami spicigeri, crassi, usque ad ultra 10 cm. diam. Poliorum rhacliis 230-370 cm. long., 
0*2 cm. diam., rigida, subtetragona, intervallis 2-0-3'5 cm. long, folioligera. Poliola 
12-0-16'0 cm. long, (rare usque ad S'O-IO'O cm. miniata), 4-5-6-0 cm. lat., firme 
membranacea, subtus pallidiora, petiolis crassiusculis in sicco rugatis 0-3-0-6 cm. 
long, insidentia ; costse secundarise utrinque circa 8, plerumque angulis latis insertge, 
aperte arcuata?. Spicse pen dulse, usque ad 27 '0 cm. long, (exstant nonnullte vero 
ssepe 10'0-160 cm. long.), 1-5-2-0 cm. diam., rigidse, minute pubescentes, ima basi 
incrassatse. Bractese nobis baud suppetunt. Plores pedicellis brevibus, crassiusculis, 
subtetragonis, 0'1-0*15 cm. long, fulti. Calycis lobi orati, obtusi, puberuli, circa 
0*1 cm. long. Petala calycem paullo excedentia. Ovarium oblongo-ovoideum, sur- 
sum angustatum, arete tomentosum, 0-25 cm. long. Stigmata ligulata, obtusa, 
0"004 cm. long. Baccse oblongo-ovoidese, 0"8 cm. long., 0*6 cm. diam., primo 
coccinese, deinde nigrse. 
In general appearance this plant resembles a Picramnia, but tbe gland-dotted leaves 

are peculiar, and may indicate a Hutaceous affinity. Unfortunately, mature flowers are 

not present, bence tbe impossibility of definitely placing tbe plant. 

OCHNACEiE. 

OuRATEA PURPTJEIPES (sp. nov.). Eruticosa, ramulis gracilibus, teretibus, glabris, cortice 
papyraceo brunescente vel cinerascenteobtectis; foliis mediocribus, tenuiter coriaceis, 
glaberrimis, oblanceolatis vel oblanceolato-oblongis, basi coarctatis, margine serrulatis, 
petiolis brevibus crassiusculis insidentibus ; costa media utrinque prominente, costis 
secundariis j)arum conspicuis, utrinque circa 10-12, aperte arcuatis ; venulis pluribus 
borizontalibus, raro dichotomis ; paniculis terminalibus axillaribusve, quam folia 
brevioribus, racemiformibus, paucifloris, eorum ramulis verisimiliter ebracteatis, 
fuscis, minute pub escentibus; pedicellis gracilibus, alabastra ovoidea acutiuscula sub- 
aequantibus; sepalis inaequalibus, oblongis vel ovato-oblongis; petalis obovato-oblono-is, 
basi late cuneatis, antlieras bene excedentibus; antheris lineari-oblono-is, breviter acu- 
minatis, rugulosis ; gynopboro ovario sequilongo ; stylo incrassato, ovarium 4-plo 
excedente ; drupa solitaria (an semper ?), anguste ovoidea, gynopborum ovoideum 
triplo excedente. 
Sah. Reperi in ripa fl. Paraguay juxta Santa Cruz mens. Oct. florentem et fructi- 
ficantem. (IS". 512.) 

Bamuli O'l-OS cm. diam., intervallis plerumque I'O-S'O cm. long, foliiireri. Folia 
pleraque 7-0-10-5 cm. long., 2-0-3-5 cm. lat., nitidula, subtus pallidiora, creberrime 
ac minutissime pellucid o-punctat a, margine breviter cartilaginea, basi Integra; 
petioli crassiusculi, supeme late excavati, dorso transverse rugati, 0'3 cm. Ion"*. 
StipulsB ovato-lanceolatse, acuminatee, 0-35 cm. long. Paniculte circa 4-0 cm. Ion"-, 
basi semel ramosse vel ad meros racemos reductse. Pedicelli sa^pe bini vel terni, 
ascendentes, sub fllore 0-4 cm. long. Alabastra 0-5 cm. long. Sepala 0-4-0-45 
cm. long., 0- 2-0- 25 cm. lat., margine nunc late, nunc vix scariosa, Petala vix 
omnino matura 05 cm. long., basi 017, superne 0-3 cm. lat. Antberee 0-3-0-35 



OF THE MxlTTO GROSSO EXPEDITION, lSOl-92. 333 

cm. long., basi 0'5 cm. lat., apice ipso flavescentes, ceteroquin aurantiaco-rubescentes. 

Gynopbomm pentagonum, 0'06 cm. long., sub fructu purpureus, 03 cm. long, ct 

0-25 cm. diam. Stylus 0-21 cm. long. Dmpic vix 0-9 cm. long., 0'5 cm. diam., 

basi ipsa breviter attenuatae, politae, vivaB virides. 
Speciei nuUi mihi cognitaB affinis, et imprimis venatione, foliis mediocribus utrinque 
glaberrimis, tenuiter coriaceis, margine serrulatis, inflorescentia racemiformi, pedicellis et 
alabastris sequilongis facili negotio recognoscenda. 

Huic affinis, ab ea racemorum et pedicellorum indiimento discrejians, petalis majori- 
bus, antheris longioribus, drupis parvis binis gynophoro subgloboso roseo insidentibus, 
exstat insequens : — 

OuRATEA nosiPES (sp. nov.). Pruticosa, ramulis sat validis, teretibus, cortice crustaceo- 
cinereo obductis; foliis glaberrimis, breviter petiolatis, tenuiter coriaceis, oblongis vel 
oblongo-oblanceolatis, obtusis, deorsum cuneatim angustatis, margine infcrneintegris, 
superne serrulatis, costa media utrinque prominente, costis secundariis aegre aspec- 
tabilibus intrusis latere circa 12 aperte arcuatis, demum margini parallelis, interjectis 
costis pluribus tertii ordinis subparallelis, crebre ramosis ; inflorescentiA. racemosj\ (an 
semper ?), axillari vel terminali, foliis breviore, basi bracteata paucifloriV, fulvo- 
tomentosa, dein puberula, pedicellis fulvo-tomentosis, alabastris ol)longis obtusis 
sequilongis ; sepalis ina^qualibus, oblongis vel ovato-oblongis, plus minus scariose 
marginatis ; petalis obovatis, basi late cuneatis, antheras suba;quantibus ; antheris 
linearibus, acuniinatis, rugulatis ; gynophoro ovario vix sequilongo, stylo fiiifornii, 
ovarium triplo excedente; drupis binis ovoideis, gynopliorum subglobosum duplo 
excedentibus. 
H.ab. Ad Santa Cruz mens. Oct. floret et fructificat. (N. 364.) 

Erutex diffusus, ultrametralis. Ramuli erecti, usque ad 0*4 cm. diam. Folia 7'5-ll*0 
cm. long., 2-0-3'5 cm. lat., utrinque nitidula, subtus parum decoloria, minute 
pellucido-punctata, margine breviter cartilaginea ; petioli 02 cm. long., superne 
excavati, dorso rugati. Stipulae firmse, lanceolatae, 0*2-0' 25 cm. long., in sicco 
fuscae. Kacemi usque 3*0 cm. long., bractese stipulis similes. Pedicelli et alabastra 
O'i cm. long., hsec basi 0*22 cm. lat. Sepala Oo cm. long., 0-2-0"3 cm. lat., 
apice obtusa vel acutiuscula, dorso minute pubescentia. Petala inasqualia, 0"6-0'7 
cm. long., ima basi 0*12 cm. lat., superne usque 0*4 cm. lat. Antherse 0-5-0-6 cm. 
long., in sicco, apice flavescente exempto, fusco-bruneae. GjTiophorum sub- 
pentagonum 008 cm. long., sub fructu roseum, 0'3 cm. long. Ovarium 
01 cm. long., 0*13 cm. diam. Stylus 0'32 cm. long. Drupse 0-5-0*7 cm. long., 
0-4_0-5 cm. diam., nitidulae, vivse virides. 

OuiiATEA RiEDELiANA, Engl, in Mart. Fl. Bras. xii. pars ii. p. 322. 

Kah. Ad Cuyaba satis abundat, reperi itaque ad Santa Cruz. Floret menss. Sept.- 
Nov. (Nn. 24, 474.) 

The Santa Cruz plant has larger and more deeply serrated leaves tlian tbose of the 
type : the leaves attain a length of 14 centimetres ; and the nerves also are more plainly 
marked. [Bras. Or.] 



334 ME. SPENCER LE M. MOOEE— PHANEROGAMIC BOTANY 

OUKATEA OEGYALTS (sp. iiov.). Erecta, glaberrima, ramulis subtetragonis, corticc cinereo 

obtectis; foliis coriaceo-raembranaceis, oblongo-oblanccolatis, obtusissimis vel obtusis, 

basi acutis, margine (parte infima Integra exempta) impariter spinuloso-serrulatis, 

petiolis brevibiis fultis ; cost is secundariis utrinqiie circa 10, plus minus fornicatis, 

demiim margini parallelis, supra eminentibus, venulis perplurimis rarius anastonio- 

santibus, magna pro parte costis secundariis perpendicularibus vix prominulis ; 

stipulis parvis lineari-lanceolatis acuminatis; panicula terminali anguste tliyrsoidea, 

ejus ramulis superioribus abbreviatis, ramis omnibus subsimplicibus et paucifloris ; 

pedicellis tenuibus, alabastra paullulum superantibus ; petalis late obovatis, dimidio 

inferiore gradatim cuneatis ; antheris attenuatis, rugulosis, insequilongis, longioribus, 

quam petala paullo brevioribus ; ovario 5-gyno, gynophoro brevi insidente ; carpidiis 

oYoideis ; stylo superne sensim angustato, antheras baud superante. 

Hah. Ad Santa Cruz floret mens. Nov. (N. 455.) 

Prutex prolixus, erectus, orgyalis, caule graciJi, apice solummodo foliigero. Ramuli 

pauci, ascendentes, 0'2-0'3 cm. diam. Polia et flores iisdem temporibus geniti, ilia 

12*0 cm. long., vix 4"0 cm. lat., supra nitidula, subtus pallidiora ; petioli 

0*3 cm. long., crassiusculi, supra canaliculati ; stipulae 0'8 cm. long., crebre nervosae. 

Panicula 140 cm. long., inferno circa 4*0 cm. lat., mox usque ad 2*0 cm. attenuata. 

Pedicelli 08 cm. long. Alabastra anguste ovoidea, acuta, 0*6 cm. long., 0-3 cm. lat. 

Sepala oblonga, sub apice breviter mucronulata, 0*75 cm. long., exteriora crassi- 

uscula, interiora margine scariosa. Petala 0-75 cm. long., summum 0*6